search

liruilong940607 / glcic.pytorch

This is a implement of the Siggraph2017 paper: "Globally and Locally Consistent Image Completion"
15 stars 7 forks source link

Another bug for data #2

Open miaojiang1987 opened 5 years ago

miaojiang1987 commented 5 years ago

Hi Ruilong Li,

I found another bug after modifying epochs for G and D.

Here is the report. Traceback (most recent call last): File "train.py", line 592, in main() File "train.py", line 531, in main train(dataLoader, model_G, model_D, epoch) File "train.py", line 450, in train losses_G.update(loss_G.data[0], input3ch.size(0)) AttributeError: 'AverageMeter' object has no attribute 'data'

Does the data here mean val?

Best, Miao

liruilong940607 commented 5 years ago

@miaojiang1987 Sorry for the delay and thank you for pointing this bug.

I've fixed it and update the repo by modifying Line 449. in train.py.

muni2773 commented 5 years ago

@miaojiang1987 Hi Miao, Can you please point me towards the dataset you used for train, validation and mask images. Appreciate your help. Thanks.

miaojiang1987 commented 5 years ago

@miaojiang1987 Hi Miao, Can you please point me towards the dataset you used for train, validation and mask images. Appreciate your help. Thanks.

@muni2773 , Hi, sorry that those dataset is private and belongs to an institution, I can not share. In addition to that, I also have trained places dataset.

muni2773 commented 5 years ago

@miaojiang1987 Thanks Miao - I downloaded the Places dataset and used the small images Test as Train and Val as Val. I did 6 epochs with pkls in the snapshot directory but when I run is_train=false it is looking for a mask dataset. Do you know where the mask dataset is for Places?