Open lgvaz opened 4 years ago
When doing the "increase resolution as training proceeded trick", we can use the output of the just finished trained model as the input of the new one (instead of the original image).
When doing the "increase resolution as training proceeded trick", we can use the output of the just finished trained model as the input of the new one (instead of the original image).