desy-ml/cheetah
Fast and differentiable particle accelerator optics simulation for reinforcement learning and optimisation applications.
This project helps particle accelerator physicists and engineers quickly simulate how beams behave within accelerator structures. You input your accelerator design (lattice) and beam parameters, and it outputs detailed beam dynamics, allowing for rapid iteration and optimization. It's designed for those working on tuning, identifying system properties, or integrating machine learning into accelerator control.
Use this if you need fast, precise simulations of particle beam dynamics for tasks like accelerator tuning, system identification, or integrating machine learning to optimize beam paths.
Not ideal if you're a casual user looking for a simple, non-differentiable beam simulation tool or if you do not work with particle accelerators.
Stars
63
Forks
25
Language
Python
License
GPL-3.0
Category
Last pushed
Mar 12, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/desy-ml/cheetah"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
arogozhnikov/hep_ml
Machine Learning for High Energy Physics.
CompPhysics/ComputationalPhysics2
Advanced course in Computational Physics, see texbook at...
FNALLPC/machine-learning-hats
FNAL LPC Machine Learning HATS
DeepLearningForPhysicsResearchBook/deep-learning-physics
This project contains additional material for the textbook Deep Learning for Physics Research by...
iml-wg/HEPML-LivingReview
Living Review of Machine Learning for Particle Physics