The 4x generator of SRGAN https://github.com/novwaul/SRGAN
Item | Setting |
---|---|
Train Data | DIV2K |
Validation Data | DIV2K |
Test Data | Set5, Set14, Urban100 |
Scale | 4x |
Loss | L1 |
Optimizer | Adam |
Scheduler | Multistep(g=0.5, m=[0.75 x epochs, 0.90 x epochs]) |
Iterations | 2e5 |
Batch | 8 |