Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Errors when running train.py #1

Open
hugeinteger opened this issue Jul 10, 2018 · 3 comments
Open

Errors when running train.py #1

hugeinteger opened this issue Jul 10, 2018 · 3 comments

Comments

@hugeinteger
Copy link

hugeinteger commented Jul 10, 2018

  1. I guess there is a missing argument (save_entry) in arguments parsing, I've added lines below in 'train.py':
    parser.add_argument('--save_every', type=int, default=5000,
    help='How often do you want to save result?')
  2. What should be trn-dir for running train.py?

python train.py --trn_dir='data/train' --style_path='style/abstract_1.png', --lambda_s=500000

Seems it should be downloaded like below:

mkdir -p data/train
cd data/train
wget http://www.vlfeat.org/matconvnet/models/beta16/imagenet-vgg-verydeep-19.mat
wget http://msvocds.blob.core.windows.net/coco2014/train2014.zip
unzip train2014.zip

@lyupei
Copy link

lyupei commented Jul 13, 2018

yes, you should add this argument and some CODE or it will NOT save weights until the end of train.

the trn_dir should be a directory containing some image folders or it will not find any images to use, such as coco2014/train_images/cocoxxxx.jpg you are supposed to pass coco2014 to this argument.

@hugeinteger
Copy link
Author

Is it possible to train a model for a new style image on a regular machine without GPUs?

@lyupei
Copy link

lyupei commented Aug 22, 2018

It may take several weeks...

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants