API reference

This API reference contains a collection of all functions, modules, and objects available in mlip, along with detailed descriptions of their purpose and operation.

• Data processing
  • Graph
  • Single Graph Dataset Builder
  • Graph Dataset Builder
  • Chemical System Readers
  • Chemical System
  • Dataset Info
  • Dataset Configs
  • Graph Dataset
  • Combined Graph Dataset
  • Data parallelization and batch prefetching
  • Data split
  • Batching Helpers
  • Helper functions
• Models
  • MLIP Network: Base Class
  • MLIPNetworkConfig
  • MACE
  • ESEN
  • NequIP
  • ViSNet
  • Force Field predictor
  • Conservative Predictor
  • Hessian Predictor
  • Force Field
  • Inference Context
  • Blocks
  • Readout blocks
  • Prediction
  • Properties
  • Model I/O
  • Loss
  • Model options: Enum classes
• Model training
  • Training Loop
  • Training Loop Config
  • Training state
  • Training step
  • Optimizer
  • Exponential Moving Average (EMA)
  • IO handling during training
  • Model evaluation
  • Fine-tuning utilities
• Simulations
  • Simulation Engine: Interface
  • Simulation Config
  • Simulation State
  • Simulation Enums
  • Create graph from ASE atoms for JAX-MD
  • JAX-MD Simulation Engine
  • JAX-MD Simulation Config
  • ASE Simulation Engine
  • MLIP ASE Calculator
  • ASE Simulation Config
  • Temperature scheduling
  • Batched inference
  • NEB Simulation Engine
  • NEB Simulation Config
• Utilities
  • Multi-host utilities