FLOPs?

Question

vztu opened this issue 3 years ago · comments

Hi,

Thanks for this great work! Could you provide the FLOPs/MACs of your SwinIR model?

Jingyun Liang · Answer 1 · Sat Sep 11 2021 16:11:34 GMT+0800 (China Standard Time)

We have already provided the FLOPs in README.md.

FYI

Method	Training Set	Training time (8 GeForceRTX2080Ti batch=32, iter=500k)	Y-PSNR/Y-SSIM on Manga109	Run time (1 GeForceRTX2080Ti , on 256x256 LR image)*	#Params	#FLOPs	Testing memory
RCAN	DIV2K	1.6 days	31.22/0.9173	0.180s	15.6M	850.6G	593.1M
SwinIR	DIV2K	1.8 days	31.67/0.9226	0.539s	11.9M	788.6G	986.8M

wanglixilinx · Answer 2 · Mon Sep 13 2021 11:04:35 GMT+0800 (China Standard Time)

Hi, @JingyunLiang ,
The FLops in the above table are calculated with the input resolution=256x256?

Jingyun Liang · Answer 3 · Mon Sep 13 2021 14:39:32 GMT+0800 (China Standard Time)

Runtime, #FLOPs and testing memory are all calculated on 256x256 LR image for X4 SR.

Zhengzhong Tu · Answer 4 · Thu Sep 16 2021 14:34:06 GMT+0800 (China Standard Time)

Hi, @JingyunLiang, how about the FLOPs for the denoising model on 256x256 inputs?

Jingyun Liang · Answer 5 · Thu Sep 16 2021 16:26:33 GMT+0800 (China Standard Time)

It is similar to the SR model because the main difference is the upsample module. The exact FLOPS for denoising on 256x256 models is 787.993G.

Jingyun Liang · Answer 6 · Mon Sep 27 2021 15:35:20 GMT+0800 (China Standard Time)

Feel free to open it if you have more questions.