Semantic Segmentation on RTX 3060ti is very slow

facebookresearch / playtorch

PlayTorch is a framework for rapidly creating mobile AI experiences.

https://playtorch.dev/

MIT License

829 stars 101 forks source link

Semantic Segmentation on RTX 3060ti is very slow #55

Closed Razamalik4497 closed 2 years ago

Razamalik4497 commented 2 years ago

Description

I'm using RTX 3060ti for semantic Segmentation , it takes long time to segment out object ( inference on video) 2 minutes video covering in 10 minutes.

How can I speed up my model ?

Images

Website

https://github.com/maudzung/TTNet-Real-time-Analysis-System-for-Table-Tennis-Pytorch

Credit (optional)

No response

Source Code (optional)

No response

raedle commented 2 years ago

Hi @Razamalik4497, on what platform are you running inference?

clarksandholtz commented 2 years ago

Closing due to inactivity and looks like this is a PyTorch question, not PyTorch Live