glitchgan.tf.gan_models

TensorFlow/Keras GAN model classes for cDVGAN.

Includes: - cWGAN — conditional Wasserstein GAN with gradient penalty - cDVGAN — adds a first-derivative discriminator - cDVGAN2 — adds first and second derivative discriminators - build_gan() — factory function

Module Contents

glitchgan.tf.gan_models.NUM_CLASSES = 7

class glitchgan.tf.gan_models.cWGAN(signal_length=8192, num_classes=NUM_CLASSES, noise_dim=100, d_steps=5, gp_weight=10.0, lr=0.0001)

Bases: keras.Model

Conditional Wasserstein GAN with gradient penalty.

noise_dim = 100

num_classes = 7

d_steps = 5

gp_weight = 10.0

generator

discriminator

g_optimizer

d_optimizer

train_step(data)

class glitchgan.tf.gan_models.cDVGAN(signal_length=8192, num_classes=NUM_CLASSES, noise_dim=100, d_steps=5, gp_weight=10.0, lr=0.0001)

Bases: keras.Model

Conditional Dual-discriminator Variational GAN (first derivative).

noise_dim = 100

num_classes = 7

d_steps = 5

gp_weight = 10.0

generator

discriminator

deriv_discriminator

g_optimizer

d_optimizer

d2d_optimizer

train_step(data)

class glitchgan.tf.gan_models.cDVGAN2(signal_length=8192, num_classes=NUM_CLASSES, noise_dim=100, d_steps=5, gp_weight=10.0, lr=0.0001)

Bases: keras.Model

cDVGAN with an additional second-derivative discriminator.

noise_dim = 100

num_classes = 7

d_steps = 5

gp_weight = 10.0

generator

discriminator

deriv_discriminator

deriv2_discriminator

g_optimizer

d_optimizer

d2d_optimizer

d2d2_optimizer

train_step(data)

class glitchgan.tf.gan_models.GlitchGAN(noise_dim=100, d_steps=5, gp_weight=10.0, lr=0.0001)

Bases: cDVGAN

cDVGAN trained on LIGO gravitational-wave glitch data.

Fixes the LIGO-specific defaults (signal length, number of glitch classes) so they don’t need to be passed at every call site. All architecture and training logic lives in cDVGAN.

SIGNAL_LENGTH = 8192

NUM_CLASSES = 7

glitchgan.tf.gan_models.build_gan(variant='cDVGAN', signal_length=8192, num_classes=NUM_CLASSES, noise_dim=100, d_steps=5, gp_weight=10.0, lr=0.0001)

Instantiate a TF GAN variant by name.

Parameters:

variant (str) – One of "cWGAN", "cDVGAN", "cDVGAN2".

Return type:

keras.Model