==========
tft-torch
==========

.. include:: badges.rst

``tft-torch`` is a ``Python`` library that
implements the model presented in the paper `Temporal Fusion Transformers for Interpretable Multi-horizon Time Series
Forecasting <https://arxiv.org/abs/1912.09363>`_  by Bryan Lim, Sercan O. Arik, Nicolas Loeff and Tomas Pfister.

The library is implemented using **PyTorch** framework, and it provides a complete implementation of a time-series
multi-horizon forecasting model with state-of-the-art performance on several benchmark datasets.

This library works for Python 3.7 and higher and PyTorch 1.6.0 and higher.

``tft-torch`` also provides detailed documentation and tutorials in order to help and guide users in running experiments using
the implemented model.

.. note::
    tft-torch is a python library that can be directly used by data scientists when
    imported in  ``Jupyter Notebook`` or ``Python`` projects.

For more information, refer to our
`blogpost <https://www.playtika-blog.com/playtika-ai/multi-horizon-forecasting-using-temporal-fusion-transformers-a-comprehensive-overview-part-1/>`_.