Skip to main content

Ctrl+K

xnn is under active development (pre-1.0): APIs may change between releases. Feedback and contributions are welcome.

Getting Started
How-To Guides
Examples
User Guide
Developer Guide
Background Information
API Reference

GitHub
MolSSI

Getting Started
How-To Guides
Examples
User Guide
Developer Guide
Background Information
API Reference

GitHub
MolSSI

Section Navigation

xnn.common
xnn.gnn
- xnn.gnn.featurizers
- xnn.gnn.models
xnn.cnn
- xnn.cnn.models
  - xnn.cnn.models.schnet
    
    xnn.cnn.models.schnet.SchNet
xnn.dnn
- xnn.dnn.featurizers
  - xnn.dnn.featurizers.aev
    
    xnn.dnn.featurizers.aev.AEV
  - xnn.dnn.featurizers.symmetry_functions
    
    xnn.dnn.featurizers.symmetry_functions.AngularSymmetryFunctions
    
    xnn.dnn.featurizers.symmetry_functions.RadialSymmetryFunctions
- xnn.dnn.models
xnn.ffnn
- xnn.ffnn.common
- xnn.ffnn.models
xnn.transformer
- xnn.transformer.attention
  - xnn.transformer.attention.EdgeMultiheadAttention
- xnn.transformer.featurizers
  - xnn.transformer.featurizers.radial
    
    xnn.transformer.featurizers.radial.ExpNormalSmearing
xnn.hybrid
- xnn.hybrid.models
  - xnn.hybrid.models.bamboo
    
    xnn.hybrid.models.bamboo.ELE_FACTOR
    
    xnn.hybrid.models.bamboo.DEBYE_EA
    
    xnn.hybrid.models.bamboo.BAMBOO
    
    xnn.hybrid.models.bamboo.GETLayer
  - xnn.hybrid.models.dispersion
    
    xnn.hybrid.models.dispersion.HARTREE_KCAL_MOL
    
    xnn.hybrid.models.dispersion.D3CSODispersion

API Reference
xnn.common
xnn.common.train

xnn.common.train#

Modules

`losses`	Weighted energy / force / stress loss.
`trainer`	Training loop: batch training, device selection, validation, checkpointing.

previous

xnn.common.models.registry.register_model

next

xnn.common.train.losses

xnn - MolSSI-AI

xnn - MolSSI-AI

Built with Sphinx · xnn by MolSSI.

© Copyright 2026, The Molecular Sciences Software Institute

Supported by the National Science Foundation under grant CHE-2136142.

The NSF