search

hanyoseob / pytorch-DIP

Deep Image Prior
https://github.com/hanyoseob/pytorch-DIP
1 stars 1 forks source link

ValueError: operands could not be broadcast together with shapes #1

Open askiiRobotics opened 4 years ago

askiiRobotics commented 4 years ago

Hi! Thank you for the project. I've faced with a problem.

pytorch-DIP-master\dataset.py", line 64, in getitem

ValueError: operands could not be broadcast together with shapes (188,512,3) (512,512,1)

There are input = data * mask

`---------------------------------------- PARAMETER TABLES

batch_size : 1 beta1 : 0.5 data_type : float32 dir_checkpoint : ./checkpoint dir_data : ./datasets dir_log : ./log dir_result : ./results gpu_ids : [0] loss_type : l2 lr_G : 0.1 mode : train name_data : restoration2 nch_in : 1 nch_ker : 128 nch_load : 1 nch_out : 1 network_type : hourglass norm : bnorm num_epoch : 20000 num_freq_disp : 5 num_freq_save : 10 nx_in : 512 nx_load : 512 nx_out : 512 ny_in : 512 ny_load : 512 ny_out : 512 optim : adam scope : inpainting_hourglass train_continue : on`

python -V Python 3.7.6

hanyoseob commented 4 years ago

I think, you need to change the arguments related with image size such as nx* , ny , and nch_ .