Tars.models package

Submodules

Tars.models.ae module

class Tars.models.ae.AE(q, p, n_batch=100, optimizer=<function adam>, optimizer_params={}, clip_grad=None, max_norm_constraint=None, seed=1234)

Bases: Tars.models.model.Model

test(test_set, n_batch=None, verbose=True)

train(train_set, verbose=False)

Tars.models.gan module

class Tars.models.gan.GAN(p, d, n_batch=100, p_optimizer=<function adam>, d_optimizer=<function adam>, p_optimizer_params={}, d_optimizer_params={}, p_critic=<function <lambda>>, d_critic=<function <lambda>>, p_clip_param=None, d_clip_param=None, p_clip_grad=None, d_clip_grad=None, p_max_norm_constraint=None, d_max_norm_constraint=None, l1_lambda=0, seed=1234)

Bases: Tars.models.model.Model

gan_test(test_set, n_batch=None, verbose=False)

train(train_set, freq=1, verbose=False)

Tars.models.model module

class Tars.models.model.Model(n_batch=100, seed=1234)

Bases: object

set_seed(seed=1234)

train()

Tars.models.vae module

class Tars.models.vae.VAE(q, p, prior=None, n_batch=100, optimizer=<function adam>, optimizer_params={}, clip_grad=None, max_norm_constraint=None, train_iw=False, test_iw=True, iw_alpha=0, seed=1234)

Bases: Tars.models.model.Model

test(test_set, l=1, k=1, n_batch=None, verbose=True)

train(train_set, l=1, k=1, annealing_beta=1, verbose=False)

Module contents

class Tars.models.AE(q, p, n_batch=100, optimizer=<function adam>, optimizer_params={}, clip_grad=None, max_norm_constraint=None, seed=1234)

Bases: Tars.models.model.Model

test(test_set, n_batch=None, verbose=True)

train(train_set, verbose=False)

class Tars.models.VAE(q, p, prior=None, n_batch=100, optimizer=<function adam>, optimizer_params={}, clip_grad=None, max_norm_constraint=None, train_iw=False, test_iw=True, iw_alpha=0, seed=1234)

Bases: Tars.models.model.Model

test(test_set, l=1, k=1, n_batch=None, verbose=True)

train(train_set, l=1, k=1, annealing_beta=1, verbose=False)

class Tars.models.GAN(p, d, n_batch=100, p_optimizer=<function adam>, d_optimizer=<function adam>, p_optimizer_params={}, d_optimizer_params={}, p_critic=<function <lambda>>, d_critic=<function <lambda>>, p_clip_param=None, d_clip_param=None, p_clip_grad=None, d_clip_grad=None, p_max_norm_constraint=None, d_max_norm_constraint=None, l1_lambda=0, seed=1234)

Bases: Tars.models.model.Model

gan_test(test_set, n_batch=None, verbose=False)

train(train_set, freq=1, verbose=False)

class Tars.models.JMVAE(q, p, prior=None, n_batch=100, optimizer=<function adam>, optimizer_params={}, clip_grad=None, max_norm_constraint=None, train_iw=False, test_iw=True, seed=1234)

Bases: Tars.models.vae.VAE

test(test_set, l=1, k=1, index=[0], sampling_n=1, missing_resample=False, type_p='joint', missing=False, n_batch=None, verbose=True)

class Tars.models.JMVAE_KL(q, p, pseudo_q, prior=None, gamma=1, n_batch=100, optimizer=<function adam>, optimizer_params={}, clip_grad=None, max_norm_constraint=None, test_iw=True, seed=1234)

Bases: Tars.models.jmvae.JMVAE

class Tars.models.CMMA(q, p, n_batch=100, optimizer=<function adam>, optimizer_params={}, clip_grad=None, max_norm_constraint=None, train_iw=False, test_iw=True, iw_alpha=0, seed=1234)

Bases: Tars.models.vae.VAE

test(test_set, l=1, k=1, type_p='normal', missing=False, n_batch=None, verbose=True)

class Tars.models.CVAE(q, p, prior=None, n_batch=100, optimizer=<function adam>, optimizer_params={}, clip_grad=None, max_norm_constraint=None, train_iw=False, test_iw=True, iw_alpha=0, seed=1234)

Bases: Tars.models.vae.VAE

test(test_set, l=1, k=1, n_batch=None, verbose=True, type_p='normal', missing=False)