Training questions

[tzayuan]

In the readme, the author mentions "For general image restoration, we first train both the stage1 and stage2 models under codeformer degradation to enhance the generating capacity of the stage2 model. In order to improve the ability for degradation removal, we train another stage1 model under Real ESRGAN degradation and utilize it during inference."

The model used in the first stage is Swinir, not Real-esrgan. Is my understanding incorrect? I urgently hope you can tell me more details about training, thank you!