Generate Faces

Have all project files been included with the submission?

The project submission contains the project notebook, called “dlnd_face_generation.ipynb”.

Have all the unit tests in the project passed?

All the unit tests in project have passed.

Has get_dataloader been implemented?

The function get_dataloader should transform image data into resized, Tensor image types and return a DataLoader that batches all the training data into an appropriate size.

Has the scale function been implemented?

Pre-process the images by creating a scale function that scales images into a given pixel range. This function should be used later, in the training loop.

Does the discriminator discriminate between real and fake mages?

The Discriminator class is implemented correctly; it outputs one value that will determine whether an image is real or fake.

Does the generator generate fake mages?

The Generator class is implemented correctly; it outputs an image of the same shape as the processed training data.

Is the weight initialization function implemented correctly?

This function should initialize the weights of any convolutional or linear layer with weights taken from a normal distribution with a mean = 0 and standard deviation = 0.02.

Are the real_loss and fake_loss functions implemented correctly?

The loss functions take in the outputs from a discriminator and return the real or fake loss.

Are appropriate optimizers defined for the networks?

There are optimizers for updating the weights of the discriminator and generator. These optimizers should have appropriate hyperparameters.

Are all adversarial networks trained correctly?

Real training images should be scaled appropriately. The training loop should alternate between training the discriminator and generator networks.

Do all models and optimizers have reasonable hyperparameters?

There is not an exact answer here, but the models should be deep enough to recognize facial features and the optimizers should have parameters that help wth model convergence.

Does the project generate realistic faces?

The project generates realistic faces. It should be obvious that generated sample images look like faces.

How could your model improve?

The question about model improvement is answered.