Generate and simulate THRML code for JAX-based Ising and Boltzmann machines.
npx skills add https://github.com/extropic-ai/thrml-skill --skill thrml
Please help me install this Agent Skill. Skill: thrml Source: https://github.com/extropic-ai/thrml-skill/tree/main Command: npx skills add https://github.com/extropic-ai/thrml-skill --skill thrml
This Skill streamlines the creation and simulation of THRML code, enabling users to efficiently build and manipulate JAX-based probabilistic models for complex sampling and training tasks.
Run the 'thrml' skill with the command: /thrml create IsingMachine --nodes 100 --edges 200 --biases normal(0, 1) --weights normal(0, 1) --beta 1.0
High-intent search queries and answers about installing and using this skill.
To generate and simulate probabilistic graphical models using JAX, you can automate the creation of THRML code for Ising and Boltzmann machines with configurable nodes, weights, and biases. This enables efficient block Gibbs sampling and training.
Block Gibbs sampling for Boltzmann machines is a probabilistic simulation technique used to train models. You configure it by specifying node structures, edge counts, weight distributions like normal(0, 1), and a sampling schedule for your JAX environment.
Yes, JAX supports automatic differentiation and vectorized operations on large graphs for Ising machines. The Skill requires JAX specifically to enable these efficient computations during model training and simulation.
To create an Ising machine model with custom weights and biases, run a command specifying the node and edge counts, parameter distributions such as normal(0, 1) for biases and weights, and the beta value for your sampling schedule.
Training probabilistic models with this approach requires the JAX library for automatic differentiation, along with networkx for graph structures and optax for optimization, to support vectorized operations on complex graphs.
Yes, this approach supports vectorized operations for large-scale probabilistic sampling. By leveraging JAX, it efficiently handles vectorized operations and automatic differentiation on large graphs during Ising and Boltzmann machine simulations.
K-Dense-AIOptimize quantum circuits with automatic differentiation across simulators and hardware.
gracefullightConvert and deploy TensorFlow models using SavedModel and TensorFlow Lite.
RamanEbrahimiConstructs and analyzes Bayesian models using PyMC.
unitarylabEnable differentiable quantum circuits and hybrid quantum-classical workflows in Python.
