How I Built a Generative AI Model that can Generate Novel Small Molecules for Drug Discovery! ??? Part 3: Putting it all Together in a WGAN

Here, we just extend keras.Model and initialize our WGAN with a generator, discriminator, and related parameters.

  • generator and discriminator are the models we defined earlier to use in our GAN.
  • discriminator_steps and generator_steps represent the number of steps to train the discriminator and generator in each iteration.
  • gp_weight is the weight assigned to the gradient penalty term.

Website