Not changing the speed by changing the amount of data

sanghyun-son / EDSR-PyTorch

PyTorch version of the paper 'Enhanced Deep Residual Networks for Single Image Super-Resolution' (CVPRW 2017)

MIT License

2.41k stars 666 forks source link

Not changing the speed by changing the amount of data #353

Open rezraz1 opened 7 months ago

rezraz1 commented 7 months ago

Hello. Why doesn't the EDSR training speed change much when I change the number of data to a significant number? In which part of the code can I change this limit? Thanks for your help. @HolmesShuan @sanghyun-son @tabetomo @5496gaoziya @mylifeasazucchini @Senwang98