Generative Adversarial Networks

astonzhang · September 17, 2020, 5:09am

https://d2l.ai/chapter_generative-adversarial-networks/gan.html

==
The PyTorch adaptation of this section was initially contributed by @StevenJokess, as reviewed and revised by @anirudh. The PR may not show up on Git due to the suspended account of the former.

StevenJokess · December 29, 2020, 10:37am

Still suck at GAN to tensorflow…
Anyone helps?

github.com

StevenJokess/d2l-en-read/blob/moreme/chapter-generative-adversarial-networks/gan-tf6.ipynb

{
 "metadata": {
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.8-final"
  },
  "orig_nbformat": 2,
  "kernelspec": {
   "name": "python37864bittf2conda1de394c9651a4d19aa25641a40b8d2d8",
   "display_name": "Python 3.7.8 64-bit ('tf2': conda)",
   "language": "python"
  }

This file has been truncated. show original

StevenJokess · April 11, 2021, 3:34pm

https://www.heywhale.com/mw/project/6062924694a58b00178fc4d6

StevenJokess · April 11, 2021, 4:23pm

TF finished!
http://preview.d2l.ai/d2l-en/PR-1716/chapter_generative-adversarial-networks/gan.html

github.com

StevenJokess/d2l-en-read/blob/moreme/chapter-generative-adversarial-networks/gan-tf-ok/gan_tf6 (3).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.8-final"
    },
    "orig_nbformat": 2,
    "kernelspec": {
      "name": "python37864bittf2conda1de394c9651a4d19aa25641a40b8d2d8",
      "display_name": "Python 3.7.8 64-bit ('tf2': conda)",

This file has been truncated. show original

telegnosis · June 1, 2021, 3:34am

loss = nn.BCEWithLogitsLoss(reduction='sum')

With this loss where reduction is ‘sum’, I think the model does not consider the data size(batch size) in gradient descent.

Isn’t it better to use
loss = nn.BCEWithLogitsLoss(reduction='mean') with
metric.add(update_D(X, Z, net_D, net_G, loss, trainer_D)* batch_size ,update_G(Z, net_D, net_G, loss, trainer_G)* batch_size, batch_size)

If it is right, Could I commit this change?

Thanks

telegnosis · June 2, 2021, 3:56am

@goldpiggy @astonzhang Thanks in advance

Liam · July 11, 2021, 1:08pm

If the generator does a perfect job, then 𝐷(𝐱′)≈1D(x′)≈1 so the above loss near 0, which results the gradients are too small to make a good progress for the discriminator. So commonly we minimize the following loss:
…
which is just feed 𝐱′=𝐺(𝐳)x′=G(z) into the discriminator but giving label 𝑦=1y=1.

The sentences above look quite confusing to me… There might be some grammatical errors in there…
Could you please rewrite these so they look coherent and natural?

Todd_Northward · July 18, 2022, 3:34am

This sentence also confuses me a lot. Authors, please, correct this. Or at least specify the source of this sentence?

zincyxsnow · October 3, 2022, 7:39am

Same to me, I don’t understand why the above loss is near 0 and the gradients become small?