Mcmc

Module: generative_models.models.energy.mcmc

Source: generative_models/models/energy/mcmc.py

Overview

MCMC sampling utilities for energy-based models.

This module provides sampling algorithms for energy-based models, including Langevin dynamics and buffer-based sampling strategies.

Classes

SampleBuffer

class SampleBuffer

Functions

init

def __init__()

energy_single

def energy_single()

energy_single

def energy_single()

energy_single

def energy_single()

improved_langevin_dynamics

def improved_langevin_dynamics()

langevin_dynamics

def langevin_dynamics()

langevin_dynamics_with_trajectory

def langevin_dynamics_with_trajectory()

persistent_contrastive_divergence

def persistent_contrastive_divergence()

sample_initial

def sample_initial()

update_buffer

def update_buffer()

Module Statistics

• Classes: 1
• Functions: 10
• Imports: 4