TensorFlow Decision Forests (TF-DF) is a library for training and serving decision forest models such as Random Forests and Gradient Boosted Trees within the TensorFlow ecosystem. It supports classification, regression, and ranking tasks through a Keras-integrated API. The library accepts Pandas dataframes converted to TensorFlow datasets, exports models as TensorFlow SavedModels, and is powered by the underlying Yggdrasil Decision Forests C++ implementation.