Reader small image

You're reading from  Generative Adversarial Networks Projects

Product typeBook
Published inJan 2019
Reading LevelIntermediate
PublisherPackt
ISBN-139781789136678
Edition1st Edition
Languages
Right arrow
Author (1)
Kailash Ahirwar
Kailash Ahirwar
author image
Kailash Ahirwar

Kailash Ahirwar is a machine learning and deep learning enthusiast. He has worked in many areas of Artificial Intelligence (AI), ranging from natural language processing and computer vision to generative modeling using GANs. He is a co-founder and CTO of Mate Labs. He uses GANs to build different models, such as turning paintings into photos and controlling deep image synthesis with texture patches. He is super optimistic about AGI and believes that AI is going to be the workhorse of human evolution.
Read more about Kailash Ahirwar

Right arrow

Hyperparameter optimization

The model that we trained might not be a perfect model, but we can optimize the hyperparameters to improve it. There are many hyperparameters in a 3D-GAN that can be optimized. These include the following:

  • Batch size: Experiment with values of 8, 16, 32, 54, or 128 for the batch size.
  • The number of epochs: Experiment with 100 epochs and gradually increase it to 1,000-5,000.
  • Learning rate: This is the most important hyperparameter. Experiment with 0.1, 0.001, 0.0001, and other small learning rates.
  • Activation functions in different layers of the generator and the discriminator network: Experiment with sigmoid, tanh, ReLU, LeakyReLU, ELU, SeLU, and other activation functions.
  • The optimization algorithm: Experiment with Adam, SGD, Adadelta, RMSProp, and other optimizers available in the Keras framework.
  • Loss functions: Binary cross entropy is the loss...
lock icon
The rest of the page is locked
Previous PageNext Page
You have been reading a chapter from
Generative Adversarial Networks Projects
Published in: Jan 2019Publisher: PacktISBN-13: 9781789136678

Author (1)

author image
Kailash Ahirwar

Kailash Ahirwar is a machine learning and deep learning enthusiast. He has worked in many areas of Artificial Intelligence (AI), ranging from natural language processing and computer vision to generative modeling using GANs. He is a co-founder and CTO of Mate Labs. He uses GANs to build different models, such as turning paintings into photos and controlling deep image synthesis with texture patches. He is super optimistic about AGI and believes that AI is going to be the workhorse of human evolution.
Read more about Kailash Ahirwar