search

speedinghzl / pytorch-segmentation-toolbox

PyTorch Implementations for DeeplabV3 and PSPNet
MIT License
768 stars 167 forks source link

Reimplement results #11

Closed LeiyuanMa closed 5 years ago

LeiyuanMa commented 5 years ago

Due to GPU limitation,I set the BS=2,STEPS= 160000 and GPU_IDS= 0,(also Titan XP),But I'm confused that the my reimplement results is not so well,mIOU is only 0.687...could you double check are there some details go wrong?As I found the args.ohem was set to false,the set_bn_eval and set_bn_momentum function seems not working...As for evaluate,It seems that args.recurrence working in any way...etc..I'm not sure if these will influence the reimplement results or not,could you give me some advice?

speedinghzl commented 5 years ago

Please refer to ReadMe! This repo requires 4 x 12g GPUs because the large batch size is important for high performance.

LeiyuanMa commented 5 years ago

Thanks,I tried with 2 gpu,the performance reach 0.756..

speedinghzl commented 5 years ago

Ok. It's a reasonable result. You could achieve higher performance if using 4 GPUs.