fjhzhixi / 3D-SPS

MIT License
54 stars 1 forks source link

load the pred_weight error #12

Closed xuxiaoxxxx closed 1 year ago

xuxiaoxxxx commented 1 year ago

When I load the pre_trained model, I meet this error:

RuntimeError: Error(s) in loading state_dict for VGNet:
        size mismatch for object_proj_layer.weight: copying a param with shape torch.Size([256, 288, 1]) from checkpoint, the shape in current model is torch.Size([384, 288, 1]).
        size mismatch for object_proj_layer.bias: copying a param with shape torch.Size([256]) from checkpoint, the shape in current model is torch.Size([384]).

Do I need to change the channel?