Package overview¶

.
├── .github                 # CI workflows and repository templates
├── Makefile                # Common development, build, and test commands
├── docs                    # Documentation sources and generated site assets
│   └── agents              # Agent reference documentation
├── examples                # Runnable scripts for training and benchmarking
│   ├── hgnn.py             # HGNN training and benchmarking example
│   ├── ...                 # Additional model examples
│   └── villain.py          # VilLain training and benchmarking example
├── hypertorch              # Main HyperTorch Python package
│   ├── data                # Datasets, data loaders, and sampling
│   ├── hlp                 # Hyperlink prediction pipelines
│   ├── integration_tests   # Integration test suite
│   ├── models              # Model implementations
│   ├── nn                  # Neural-network layers and reusable modules
│   ├── tests               # Unit test suite
│   ├── train               # Training and benchmarking logic, and loggers
│   ├── types               # Shared type definitions
│   └── utils               # General-purpose utility functions
├── hypertorch_logs         # Local logs and outputs from experiments
│   └── experiment_0        # Outputs produced by experiment_0
│       ├── comparison      # Cross-model metric comparisons
│       ├── hgnn            # HGNN-specific experiment results
│       ├── ...             # Additional model-specific outputs
│       └── villain         # VilLain-specific experiment results
├── pyproject.toml          # Metadata, dependencies, and build settings
├── uv.lock                 # Lockfile for dependencies
└── zensical.toml           # Documentation build and site configuration