Size mismatch when loading the pretrained pytorch model.

[xinyusun1997]

Below is the error message:
"size mismatch for ref_score_subtract.weight: copying a param with shape torch.Size([1]) from checkpoint, the shape in current model is torch.Size([1, 1])."

Need to manually adjust the size of 'ref_score_subtract.weight'.

[NagabhushanSN95]

I was able to fix it as follows

state_dict = torch.load(weights_path, map_location='cpu')
state_dict['ref_score_subtract.weight'] = state_dict['ref_score_subtract.weight'].view((1,1))
self.PieAPP_net.load_state_dict(state_dict)

[shshojaei]

I was able to fix it as follows

state_dict = torch.load(weights_path, map_location='cpu')
state_dict['ref_score_subtract.weight'] = state_dict['ref_score_subtract.weight'].view((1,1))
self.PieAPP_net.load_state_dict(state_dict)

Hi, i have a question if you or @prashnani know, what's the max and min value of pieApp in two reference and generated images? I've got -0.0334 for one of my experiments, is it ok? (i use pyiqa library for calculate pieApp because it's simple to use.)