Any requirement for my own dataset?

carpedm20 / DCGAN-tensorflow

A tensorflow implementation of "Deep Convolutional Generative Adversarial Networks"

http://carpedm20.github.io/faces/

MIT License

7.15k stars 2.63k forks source link

Any requirement for my own dataset? #212

Open KiyoshiKAWASAKI opened 7 years ago

KiyoshiKAWASAKI commented 7 years ago

I tried to use my own dataset; it can start training but only 16/1089 images works. I also tried to change the command but nothing changed. Is there any requirement for my own images? Mine are from google and all in JPG format, arbitrary size.

locha0519 commented 7 years ago

Actually all of your images worked. The idx was much more smaller since it was batch idx. It equals to the # of image divided by the batch size. batch_idxs = min(len(self.data), config.train_size) // config.batch_size ... print("Epoch: [%2d] [%4d/%4d] time: %4.4f, d_loss: %.8f, g_loss: %.8f" % (epoch, idx, batch_idxs, time.time() - start_time, errD_fake+errD_real, errG))

lzzlxxlsz commented 5 years ago

Hello,have you trained the modle with your data?and if I train my data ,the thing I have to do is running the main.py --dataset madata?