Defining Global and local Discriminator

**Defining a global Discriminator and its parameters**

For global :
Use convolution layers on a concatenation of the Masked Input and Output

For Local :
Use convolution layers on a concatenation of the Original content of the masked area and Corresponding content in the output.

You may wish to make a PatchGan Discriminator you can find some good resources [here](https://www.researchgate.net/figure/The-patchGAN-discriminator-input-of-discriminator-is-either-the-pair-of-sketch-yellow_fig2_325291567)

Start by reviewing the issue's global and local discriminator requirements and the linked PatchGAN reference. Determine how the convolution layers and parameters should differ for the two discriminators. Done means both definitions cover the specified concatenated inputs and their parameters are documented or implemented.

Written by the indexing model from the issue text.