.. This file is generated by scripts/render_model_docs.py. Do not edit by hand.

SAF-Net
=======

Overview
--------

``saf_net`` adds an intensity-oriented storm baseline to the shared ``tc.track_intensity`` evaluator.

At a Glance
-----------

.. grid:: 1 2 4 4
   :gutter: 2
   :class-container: catalog-grid

   .. grid-item-card:: Hazard Family
      :class-card: catalog-stat-card

      .. container:: catalog-stat-value

         Tropical Cyclone

      .. container:: catalog-stat-note

         Public catalog grouping used for this model.

   .. grid-item-card:: Maturity
      :class-card: catalog-stat-card

      .. container:: catalog-stat-value

         Implemented

      .. container:: catalog-stat-note

         Catalog maturity label used on the index page.

   .. grid-item-card:: Tasks
      :class-card: catalog-stat-card

      .. container:: catalog-stat-value

         1

      .. container:: catalog-stat-note

         Track + Intensity

   .. grid-item-card:: Benchmark Family
      :class-card: catalog-stat-card

      .. container:: catalog-stat-value

         :doc:`Tropical Cyclone Benchmark </benchmarks/tropical_cyclone_benchmark>`

      .. container:: catalog-stat-note

         Primary benchmark-family link used for compatible evaluation coverage.


Description
-----------

``saf_net`` adds an intensity-oriented storm baseline to the shared ``tc.track_intensity`` evaluator.

The adapter keeps full trajectory outputs so it can use the same report format as the other PyHazards storm models.

Benchmark Compatibility
-----------------------

**Primary benchmark family:** :doc:`Tropical Cyclone Benchmark </benchmarks/tropical_cyclone_benchmark>`

**Mapped benchmark ecosystems:** :doc:`TCBench Alpha </benchmarks/tcbench_alpha>`

External References
-------------------

**Paper:** `SAF-Net: A spatio-temporal deep learning method for typhoon intensity prediction <https://www.sciencedirect.com/science/article/pii/S1568494623003152>`_ | **Repo:** `Repository <https://github.com/xuguangning1218/TI_Prediction>`__

Registry Name
-------------

Primary entrypoint: ``saf_net``

Supported Tasks
---------------

- Track + Intensity

Programmatic Use
----------------

.. code-block:: python

   import torch
   from pyhazards.models import build_model

   model = build_model(name="saf_net", task="regression", input_dim=8, horizon=5)
   preds = model(torch.randn(2, 6, 8))
   print(preds.shape)

Notes
-----

- Track channels are retained so the shared storm evaluator can score all baselines consistently.