How to train with parameter `--load_size=512 --fine_size=512` ?

[binhnq94]

When i try to run: CUDA_VISIBLE_DEVICES=0 python main.py --dataset_dir=my_dataset --load_size=512 --fine_size=512, a error apear:

Traceback (most recent call last):
  File "main.py", line 53, in <module>
    tf.app.run()
  File "/home/vsocr/py3/lib/python3.5/site-packages/tensorflow/python/platform/app.py", line 48, in run
    _sys.exit(main(_sys.argv[:1] + flags_passthrough))
  File "main.py", line 49, in main
    model.train(args) if args.phase == 'train' \
  File "/home/vsocr/workplace/CycleGAN-tensorflow/model.py", line 158, in train
    feed_dict={self.real_data: batch_images, self.lr: lr})
  File "/home/vsocr/py3/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 895, in run
    run_metadata_ptr)
  File "/home/vsocr/py3/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1100, in _run
    % (np_val.shape, subfeed_t.name, str(subfeed_t.get_shape())))
ValueError: Cannot feed value of shape (1, 511, 511, 6) for Tensor 'real_A_and_B_images:0', which has shape '(?, 512, 512, 6)'

And run: CUDA_VISIBLE_DEVICES=0 python main.py --dataset_dir=2k_v1 --load_size=513 --fine_size=512, error:

    tf.app.run()
  File "/home/vsocr/py3/lib/python3.5/site-packages/tensorflow/python/platform/app.py", line 48, in run
    _sys.exit(main(_sys.argv[:1] + flags_passthrough))
  File "main.py", line 49, in main
    model.train(args) if args.phase == 'train' \
  File "/home/vsocr/workplace/CycleGAN-tensorflow/model.py", line 176, in train
    self.sample_model(args.sample_dir, epoch, idx)
  File "/home/vsocr/workplace/CycleGAN-tensorflow/model.py", line 218, in sample_model
    feed_dict={self.real_data: sample_images}
  File "/home/vsocr/py3/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 895, in run
    run_metadata_ptr)
  File "/home/vsocr/py3/lib/python3.5/site-packages/tensorflow/python/client/session.py", line 1100, in _run
    % (np_val.shape, subfeed_t.name, str(subfeed_t.get_shape())))
ValueError: Cannot feed value of shape (1, 256, 256, 6) for Tensor 'real_A_and_B_images:0', which has shape '(?, 512, 512, 6)'

My images input shape: (512, 512, 3).

It seems to be a bug in code. Can you show me right way to put parameters?

[skalinin]

as i understood it, a bug in the line https://github.com/xhujoy/CycleGAN-tensorflow/blob/master/model.py#L213 you need to pass the correct parameters into the function load_train_data

[ghost]

Hi @TechBK Have you find the way to train 512x512 images?

[Deeplearning20]

@TechBK @skalinin Hello, I am new to TensorFlow and machine learning.I have the same problem with you. How did you solve it?Is it related to my trainA trainB picture size? My A B size is different, the pixel is not 256 * 256.Here is my error:ValueError: Cannot feed value of shape (0,) for Tensor 'real_A_and_B_images:0', which has shape '(?, 256, 256, 6)'

[Deeplearning20]

I have solved this problem because the test must have photos

[cschar]

@TechBK Try setting your load_size to a higher resolution, since it is cropped later to the fine size.

e.g. load_size = 586

[giaosame]

@Deeplearning20 I met the same problem with yours. Can you tell me how did you solve this problem?

[yunks128]

@giaosam Please check if all trainA, trainB, testA, testB folders exist and not empty.

[ennemoser]

Even if I have images in trainA and trainB the error keeps appearing. Is there a certain amount of images that have to be present in the train folder?Thanks for your help.

[giaosame]

@ennemoser Please note that there must be images in testA and testB.

[ennemoser]

Thank you very much for your quick response. I also have images in testA and test B. But the error keeps appearing. I define load size and fine size exactly like above. But do I have to change any other parameter/ argument?

[Baymax0525]

@skalinin thank you . a bug in the line https://github.com/xhujoy/CycleGAN-tensorflow/blob/master/utils.py#L51： def load_train_data(image_path, load_size=286, fine_size=256, is_testing=False): @TechBK you may give there a same value as load_size and fine_size. for example：513 and 512

[lijianaiml]

I think the bug is in the line https://github.com/xhujoy/CycleGAN-tensorflow/blob/master/model.py#L213: load_train_data(batch_file, is_testing=True) the load_train_data() will use default value which is :load_size=286, fine_size=256. you can change the function this way: load_train_data(batch_file, new_load_size, new_fine_size, is_testing=True)