mac744mail

> both are fine as long as the white area in the mask covers the missing region when the input is mask+image(incomplete) Thanks. I tried to ran your model on...

> did you change the --train_image_dir and --train_image_dir in train.sh to point to your dataset? Yes I think I did it correctly for own dataset. I remember at first there...

> did you check the html file in `output/` produced during training? check the html file? I didnt run the flask web demo. I just trained the model and then...

> there should be an HTML file produced in `output` by the training code. can you check it? yes I found it. I'm attaching it here (it has been converted...

> it seems that you deleted some lines of train.sh and only run a part of the training script yes, there are several "python train.py...." in train.sh. My command was...

> they are for different stages. you'll need to use all of them for training well. thanks for your time. I'll try it

> they are for different stages. you'll need to use all of them for training btw, could you please tell me the meanings of arguments "--path_objectshape_list" and "--path_objectshape_base"? I thought...

> that's right and you don't need to modify these lines. Square masks and irregular masks are randomly synthesized in training. Object masks have been prepared if you run the...

same error

> i'v updated the code to allow finetune based on my pretrained models. > > use the following command: > > ``` > ./download/download_pretrain.sh > ``` > > ``` >...