thomasjpfan/pytorch_refinenet

Issues

• I have two questions:1) I train on my dataset for implement 2 classes when last layer sigmoid activate for binary cross entropy ,but the loss didn't convergence, 2) and the output shape is (N, C, 32, 32) when input shape is (N,C,128,128), can I unsample output to mask size（128,128） ？

  #6 opened 6 years ago by liuchuanloong
  9

• how to apply it on non-square image?

  #7 opened 5 years ago by lingligithub
  0

• ValueError: optimizing a parameter that doesn't require gradients

  #3 opened 7 years ago by cltdevelop
  3

• accuracy on Pascal VOC

  #1 opened 7 years ago by ds2268
  2

• number of channel mismatch

  #5 opened 6 years ago by 1973Blunt
  3

• one of the variables needed for gradient computation has been modified by an inplace operation

  #2 opened 7 years ago by drdlfy
  4