| Parameter | Value |
|---|---|
| batch_size | 128 |
| image_size | 64 |
| nc | 3 |
| nz | 100 |
| ngf | 64 |
| ndf | 64 |
| num_epochs | 5 |
| lr | 0.0002 |
| beta1 | 0.5 |
| ngpu | 1 |
After running DCGAN on the CelebA Dataset for 5 epochs on GPU (computationally very expensive) we got the following output images along with the Generator and Discriminator losses.
| No. of Epochs | Generator Loss | Discriminator Loss |
|---|---|---|
| 1 | 0.7894 | 1.0838 |
| 2 | 0.7277 | 1.0489 |
| 3 | 0.7796 | 0.9256 |
| 4 | 0.6330 | 1.1345 |
| 5 | 0.7519 | 1.0138 |
