thrml Wiki Documentation

7 pages available

THRML: Thermodynamic Hypergraphical Model Library

THRML (Thermodynamic Hypergraphical Model Library) is a JAX-based Python library designed for building and efficiently sampling probabilistic graphical models (PGMs). Developed by Extropic, it focuses

Getting Started

This guide will walk you through the installation of the THRML library and provide a detailed explanation of a quick example demonstrating its core functionalities. THRML is designed to facilitate the

Core Concepts

THRML (Thermodynamic Hypergraphical Model Library) is built upon a set of fundamental abstractions that collectively enable the definition, representation, and efficient sampling of probabilistic grap

API Reference: Block Management & Sampling

This page details the core classes and functions responsible for managing blocks of nodes and orchestrating the sampling process in THRML. These components are fundamental to how THRML handles probabi

API Reference: Models & Factors

This page provides a detailed API reference for the core modeling components in THRML, including factors, interactions, energy-based models (EBMs), and conditional samplers. These components are essen

Examples: Ising Models

Ising models are fundamental in statistical mechanics and machine learning, particularly as a foundational type of [Energy-Based Model (EBM)](/wiki/extropic-ai/thrml/api-models-factors#energy-based-mo

Examples: Discrete Energy-Based Models

This guide demonstrates how to build and sample a discrete Energy-Based Model (EBM) in THRML, featuring a mix of `SpinNode` (binary) and `CategoricalNode` variables. This type of mixed model is common