This is technical documentation. See also: SCM Website | SCM Helpdesk | Pricing & Free trial

Python Examples Python Examples Python Examples
  • Installation
    • Windows
    • Linux
    • MacOS
    • Package manager Install optional components
  • Tutorials
    • Getting started
    • All tutorials
  • AMS Driver
    • Geometry and System
    • Geometry Optimization
    • Molecular Dynamics
    • Nudged Elastic Band (NEB)
    • PES Exploration
    • PES Scan, Linear Transit
    • Transition State Search
    • Vibrational Spectroscopy
  • Engines
    • ADF Molecular DFT
    • BAND Periodic DFT
    • DFTB Semi-empirical
    • ForceField AMBER, UFF, APPLE&P...
    • GFNFF Force field
    • Hybrid QM/MM and mixed methods
    • ML Potential Machine Learning potentials
    • MOPAC Semi-empirical
    • Quantum ESPRESSO Periodic DFT
    • VASP Periodic DFT
    • ReaxFF Reactive force field
    • ASE engine Use any ASE calculator
  • Meso/Macro
    • COSMO-RS Thermodynamic properties of fluids
    • Bumblebee OLED device modeling
    • Microkinetics
    • Zacros Kinetic Monte Carlo
  • Tools
    • GUI Graphical User Interface
    • ParAMS Parametrization tool
    • Simple Active Learning On-the-fly training of ML
    • OLED Deposition and Properties Multiscale OLED modeling
    • Reactions Discovery Automatic reaction-discovery workflow
    • ACE Reaction Network Generation of reaction networks
    • Conformers Conformers generation
    • MD Trajectory Analysis Analysis of MD trajectories
    • ChemTraYzer2 Detect reactions from MD simulations
    • Utilities Various utility tools
  • Scripting
    • Python Scripting Examples Examples for Python scripting
    • PLAMS Python Library for Automating Molecular Simulations
    • amspython Python stack shipped with AMS
    • pyCRS Python Scripting with COSMO-RS
    • SCM Base Library Python library with core modules
    • reactmap Atom mapping between reactants and products
    • Command-line tools
    • ASE calculator
    • pyZacros Python Library Zacros (Kinetic Monte Carlo)
  • AMS2026.1
  • Other versions
  • Python Examples
  • All Examples
  • Getting Started With Python Scripting In AMS
    • Getting Started: Geometry Optimization of Water
    • ChemicalSystem: Getting Started With AMS System Blocks
    • Composing AMS Input Blocks with PLAMS Settings
    • Molecule Tools for Inspection and Manipulation
    • Visualizing Structures with PLAMS
  • Structures and Data Model
    • Building Surfaces and Slabs
    • Building Packed Systems with Packmol
    • Breaking bonds in a ring molecule and modifying distances for NEB
    • Substituting Functional Groups in Molecules
    • Balancing Reaction Equations from Formulas and Structures
    • 2D Reaction Schematics from PLAMS Molecules
    • Complete guide to storing and converting PLAMS Molecules between Python libraries and file formats
  • Jobs and Automation
    • Running Many PLAMS Jobs in Parallel
    • Logging and Job Log Files in PLAMS
    • Summarizing PLAMS Jobs with JobAnalysis
    • Helium Dimer Dissociation Curve with ADF
    • Correlation Plots for Computed Results
    • Using AMS as an ASE Calculator
    • Charged Systems with the AMS ASE Calculator
    • Engine ASE with Import and File Calculators
    • Hybrid engine UseLowestEnergy
    • Reuse ForceField Parameters From Previous Calculations
    • Constrained Geometry Optimization with AMSWorker
  • Benchmarks
    • Benchmarking Reaction Energies across AMS Engines
    • Benchmarking Reaction Energies across Basis Sets
    • Benchmarking ML Potentials for Hydrocarbon Isomerization Energies
  • Molecular Dynamics and Sampling
    • Molecular Dynamics with Python
    • ReaxFF Density Equilibration
    • Diffusion Coefficient Temperature Dependence
    • Diffusion Coefficient Supercell Dependence
    • Gasphase IR spectrum from Molecular Dynamics
    • IR Spectrum of an H2O Dimer from MD
    • i-PI Path Integral MD with AMS
    • Molecule and Bond Counts from Reactive MD
    • Extract Frames from an AMS Trajectory with PLAMS
    • Convert from XYZ/OUTCAR to RKF
    • Convert RKF Trajectory to DCD Format
    • PLUMED Biasing in AMS Molecular Dynamics
    • Basic MD Trajectory Analysis with PLAMS
    • Hydrogen Bonds from MD
    • Viscosity from MD with the Green-Kubo Formalism
  • Reaction Paths and Reactivity
    • Nudged Elastic Band (NEB) for gasphase reaction with DFTB
    • Transition State: Initial guess from PESScan
    • Automated Transition-State Workflow for a Diels-Alder Addition
    • Reaction Discovery with ReaxFF
  • Electronic Structure and Spectroscopy
    • Screening Molecules for Promising Electronic Excitations
    • Reorganization Energy Recipe
    • ADF Fragment Analysis Recipe
    • Charge Transfer Integrals with ADF
    • BAND Fragment Analysis Recipe
    • NBO Analysis with ADF and PLAMS
    • Reduction and Oxidation Potentials
    • H-H Spin-Spin Coupling with ADF
    • Franck-Condon Vibronic DOS with ADF
    • Tuning the Range Separation Parameter in ADF
  • Materials and External Engines
    • Band structure with DFTB (SCC-DFTB, GFN1-xTB)
    • Band structure with BAND (HSE06, DFT+U)
    • Band structure with Quantum ESPRESSO (PBE)
    • Quantum ESPRESSO phonon band structure and DOS for BeO
    • Quantum ESPRESSO IR and Raman spectra for BeO
    • VASP geometry optimization with PLAMS
    • Surface Energy with ReaxFF
    • In-cell adsorption energies with ReaxFF
    • BSSE correction for H2O on ZnO(10-10)
    • Simulating XRD Patterns from a CIF Structure
  • Conformers, COSMO-RS, and Property Prediction
    • Conformer Generation, Rescoring, and Filtering
    • Conformer Generation for Multiple Molecules
    • Generating COSMO-RS Compounds from XYZ Files and SMILES
    • COSMO-RS Conformer Workflows for Acetic Acid
    • ADF and COSMO-RS Workflow
    • Property Prediction from SMILES with pyCRS
  • ML Potentials, ParAMS, and Active Learning
    • ML Potentials in AMS
      • M3GNet Universal Potential: M3GNet-UP-2022
      • M3GNet Custom Model
      • NEB for Li diffusion in layered LiTiS2 (M3GNet)
    • ParAMS: Train your own ML, DFTB, ReaxFF
      • Run a ParAMSJob for Lennard-Jones
      • Import Training Data with ResultsImporter
      • Train an M3GNet Potential with ParAMS
      • ML Training with PES Scans and Geometry Optimizations
      • ParAMS for DFTB: Silicon Band Structure
      • Set Up ParAMSJob Settings
      • Convert Between ParAMS and ASE Formats
    • Active Learning: Single-Molecule Series
      • Active Learning: Single-Molecule Setup and Run
      • Active Learning: Single-Molecule Results
      • Active Learning: Single-Molecule Comparison to M3GNet-UP-2022
      • Active Learning: Single-Molecule Production Simulation
      • Active Learning: Continue with a New System
    • Active Learning: Ru/H Case Study
      • Active Learning: Ru/H Part 1, Initial Reference Scans
      • Active Learning: Ru/H Part 2, Surface PES Scans
      • Active Learning: Ru/H Part 3, Gas Snapshots
      • Active Learning: Ru/H Part 4, Initial Training
      • Active Learning: Ru/H Part 5, Molecule-Gun MD
    • Active Learning: Additional Case Studies
      • Active Learning: Liquid Water Properties
      • Active Learning: Conformer Training with CREST
      • Active Learning: Committee Uncertainties and ReactionBoost for a Gasphase Reaction
      • Active Learning: Li-Vacancy Diffusion NEB in LiTiS2
  • Python API Documentation
  • ChemicalSystem vs PLAMS Molecule
  • Contribute Your Own Scripting Example
  • Help! (Frequently Asked Questions)
  1. Documentation /
  2. Python Examples /
  3. ML Potentials, ParAMS, and Active Learning

ML Potentials, ParAMS, and Active Learning¶

This section groups examples for pretrained ML potentials, custom models, and multi-step active-learning workflows.

  • ML Potentials in AMS
  • ParAMS: Train your own ML, DFTB, ReaxFF
  • Active Learning: Single-Molecule Series
  • Active Learning: Ru/H Case Study
  • Active Learning: Additional Case Studies
Previous
Property Prediction from SMILES with pyCRS
Next
ML Potentials in AMS

Copyright Software for Chemistry and Materials