search

jiawen-zhu / ViPT

[CVPR23] Visual Prompt Multi-Modal Tracking
MIT License
256 stars 18 forks source link

Time consuming of model training #5

Closed qq1018408006 closed 1 year ago

qq1018408006 commented 1 year ago

It is a great work! According to the information provided in the paper, the model training process requires 2 NVIDA Tesla A100 GPUs. So how long will it take on those GPUs?

jiawen-zhu commented 1 year ago

Hi~, the model training takes about 6 hours.

qq1018408006 commented 1 year ago

@jiawen-zhu Thanks. It sounds much cheap.