faceparsing traning issue
leexuan0128 opened this issue · comments
Line 135 in trainer.py, generate_label( ) miss the img_size parameter?
It suppose to use self.imsize as the second parameter, just like line 114 and line 115 in trainer.py. Also line 136 should be deleted because torch.from_numpy has beed called inside the generate_label funcion. Am I right?