Pretrain detector using vgg16-bn

Hi there,

I want to train detector using vgg16-bn as a backbone in https://github.com/rowanz/neural-motifs/blob/master/models/train_detector.py
however, I faced this error:

File models/…/lib/fpn/box_utils.py", line 44, in bbox_preds
xys = prior_centers[:, :2] + prior_centers[:, 2:] * deltas[:, :2]
File “/home/user/anaconda3/lib/python3.6/site-packages/torch/tensor.py”, line 321, in mul
return self.mul(other)
RuntimeError: invalid argument 3: sizes do not match at /pytorch/torch/lib/THC/generated/…/generic/THCTensorMathPointwise.cu:351

Any help please?