tensorflow/probability

This package collects tools for probabilistic reasoning in TensorFlow. It is intended to serve as a hub for development of modeling tools, inference algorithms, useful models, and general statistical computation. Taking advantage of the TensorFlow ecosystem allows straightforward combination of probabilistic methods with deep networks, gradient-based inference via automatic differentiation, and scalability to large datasets and models via hardware acceleration (e.g., GPUs) and distributed computation. Contents of this repository should be considered under active development; interfaces may change at any time. TensorFlow Probability does not currently contain any GPU-specific code; the primary difference between these packages is that tensorflow-probability-gpu depends on a GPU-enabled version of TensorFlow.