Functions — github.com/DSL-Lab/GRBM

CodeHub MCPcopy Create free account

hub / github.com/DSL-Lab/GRBM / functions

Functions47 in github.com/DSL-Lab/GRBM

⨍ Functions 47 ◇ Types & classes 3

↓ 17 callersMethodget_var

↓ 8 callersMethodprob_h_given_v

↓ 3 callersFunctioncosine_schedule

(eta_min=0, eta_max=1, T=10)

↓ 3 callersMethodenergy_grad_param

↓ 3 callersMethodmarginal_energy

↓ 3 callersMethodprob_v_given_h

↓ 2 callersMethodGibbs_Langevin_sampling_vh

(self, v, num_steps=10,

↓ 2 callersMethodGibbs_sampling_vh

(self, v, num_steps=10, burn_in=0)

↓ 2 callersMethodLangevin_sampling_v

(self, v, num_steps=10, et

↓ 2 callersMethodenergy

↓ 2 callersMethodlog_prob

↓ 2 callersMethodmarginal_energy_grad_v

↓ 2 callersMethodsampling

(self, num_samples)

↓ 2 callersMethodset_Langevin_adjust_step

↓ 2 callersFunctionvis_2D_samples

(samples, config, tags=None)

↓ 1 callersMethodCD_grad

↓ 1 callersFunctioncreate_dataset

↓ 1 callersMethodenergy_grad_v

↓ 1 callersFunctionfig2img

↓ 1 callersFunctionget_logger

(logger_name="exp_logger")

↓ 1 callersFunctionload

(model, results_folder, epoch)

↓ 1 callersMethodlog_gaussian

log density of single (diagonal-covariance) multivariate Gaussian

↓ 1 callersMethodlog_metropolis_ratio_Gibbs_Langevin

Metropolis-Hasting ratio of accepting the move from old to new state

↓ 1 callersMethodlog_metropolis_ratio_Langevin_one_step

Metropolis-Hasting ratio of accepting the move from old to new state

↓ 1 callersMethodmarginal_energy_grad_param

↓ 1 callersMethodnegative_grad

↓ 1 callersMethodpositive_grad

↓ 1 callersMethodreconstruction

↓ 1 callersMethodreset_parameters

↓ 1 callersFunctionsave

(model, results_folder, epoch)

↓ 1 callersFunctionsave_gif_fancy

(imgs, nrow, save_name)

↓ 1 callersFunctionsetup_logging

↓ 1 callersFunctionshow_img

(matrix, title)

↓ 1 callersFunctiontrain

(model, train_loader, optimizer, config)

↓ 1 callersFunctiontrain_model

Let us train a GRBM and see how it performs

↓ 1 callersFunctionunnormalize_img_tuple

(img_tuple, mean, std)

↓ 1 callersFunctionvis_density_GMM

(model, config)

↓ 1 callersFunctionvis_density_GRBM

(model, config, epoch=None)

↓ 1 callersFunctionvisualize_sampling

(model, epoch, config, tag=None, is_show_gif=True)

Method__getitem__

(self, visible_size, hidden_size, CD_step=1,

(self, samples)

p(x) = sum_i w[i] N(mu[i], sigma[i]^2 * I) config: w: shape K X 1, mixture coefficients, must sum to 1 m

Methodlangevin_sampling

(self, x, num_steps=10, eta=1.0e+0, is_anneal=False)

(self, v_init, num_steps=1, save_gap=1)

Methodset_Langevin_eta