# Why is the discriminator necessary in a GAN?

## post by cgold on Apr 5, 2020

I am currently attempting to code a GAN to learn to generate images of hand drawn digits (MNIST dataset) and am having little success. My discriminator / generator are failing to converge and my generator output is noise.

After trying to solve the problem for a while, I tried not using the discriminator at all, and just training the generator with a desired output of a sampled image, and within minutes, the generator had learned to output very realistic images.

Why is this a bad idea? I haven’t seen any use of this method online and it is not used in GANs, however seems to be much simpler and works (for me) much better. Does it overfit too much or is there another problem?

Thanks

1.6k views

### Related topics

| Topic | Replies | Views | Activity |
| --- | --- | --- | --- |
| [About Deep Learning General (DLG)](https://deeptalk.lambda.ai/t/about-deep-learning-general-dlg/15)<br>[Deep Learning: General Discussion](https://deeptalk.lambda.ai/c/dlg/6) | [0](https://deeptalk.lambda.ai/t/about-deep-learning-general-dlg/15/1) | 2.1k | [Oct 2017](https://deeptalk.lambda.ai/t/about-deep-learning-general-dlg/15/1) |
| [Project approach consultation - RF / DL / …?](https://deeptalk.lambda.ai/t/project-approach-consultation-rf-dl/1228)<br>[Deep Learning: General Discussion](https://deeptalk.lambda.ai/c/dlg/6) | [0](https://deeptalk.lambda.ai/t/project-approach-consultation-rf-dl/1228/1) | 1.5k | [Oct 2019](https://deeptalk.lambda.ai/t/project-approach-consultation-rf-dl/1228/1) |
| [Are gamma and beta just for removing extra 1/(epsilon)^(½) when mu and sigma all equal zero? I don’t get why use z tilde in Andrew’s deep learning course?](https://deeptalk.lambda.ai/t/are-gamma-and-beta-just-for-removing-extra-1-epsilon-when-mu-and-sigma-all-equal-zero-i-don-t-get-why-use-z-tilde-in-andrew-s-deep-learning-course/1840)<br>[Deep Learning: General Discussion](https://deeptalk.lambda.ai/c/dlg/6) | [0](https://deeptalk.lambda.ai/t/are-gamma-and-beta-just-for-removing-extra-1-epsilon-when-mu-and-sigma-all-equal-zero-i-don-t-get-why-use-z-tilde-in-andrew-s-deep-learning-course/1840/1) | 1.2k | [Aug 2020](https://deeptalk.lambda.ai/t/are-gamma-and-beta-just-for-removing-extra-1-epsilon-when-mu-and-sigma-all-equal-zero-i-don-t-get-why-use-z-tilde-in-andrew-s-deep-learning-course/1840/1) |
| [Basic idea of Convolution Neutral Network](https://deeptalk.lambda.ai/t/basic-idea-of-convolution-neutral-network/1753)<br>[Technical Help](https://deeptalk.lambda.ai/c/ti/10) | [0](https://deeptalk.lambda.ai/t/basic-idea-of-convolution-neutral-network/1753/1) | 1.2k | [Jul 2020](https://deeptalk.lambda.ai/t/basic-idea-of-convolution-neutral-network/1753/1) |
| [Deep Learning Community Forum & Discord Server](https://deeptalk.lambda.ai/t/deep-learning-community-forum-discord-server/760)<br>[Deep Learning: General Discussion](https://deeptalk.lambda.ai/c/dlg/6) | [1](https://deeptalk.lambda.ai/t/deep-learning-community-forum-discord-server/760/1) | 5.1k | [Oct 2021](https://deeptalk.lambda.ai/t/deep-learning-community-forum-discord-server/760/6) |