# tensorflow/probability

Probabilistic reasoning and statistical analysis in TensorFlow

Repository: https://github.com/tensorflow/probability
Canonical: https://ross.abutalabs.com/products/probability
Homepage: https://www.tensorflow.org/probability/
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Topics: tensorflow, bayesian-methods, deep-learning, machine-learning, data-science, neural-networks, statistics, probabilistic-programming
Last push: 2026-08-06T22:18:55+00:00

## Health v2 (maintenance only)
Score: 66/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 96, release rhythm 8, longevity 100
- inputs: {"age_days": 3236, "days_push": 27, "days_rel": 663, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4427, forks 1124 (observed 2026-08-28T04:08:48.908481+00:00)

## What it is
TensorFlow Probability is a Python library for probabilistic reasoning and statistical analysis built on TensorFlow, with a JAX substrate. It provides probability distributions, bijectors, joint distribution modeling, probabilistic neural network layers, and inference tools like MCMC and variational inference.

## Use cases
- fit bayesian regression models with uncertainty estimates
- run MCMC sampling for probabilistic models
- build deep probabilistic models with probabilistic keras layers
- model distributions and transformed random variables with bijectors
- do variational inference on large datasets with GPU acceleration
- perform probabilistic programming in JAX
- fit GLMs like logistic regression in TensorFlow

## When to choose
- you need probabilistic or Bayesian modeling integrated with deep learning
- you want scalable inference on GPU/TPU hardware
- you need a rich library of distributions, bijectors, and samplers
- you work in the TensorFlow or JAX ecosystem

## When to avoid
- you only need classical statistics without deep learning integration
- your project uses PyTorch instead of TensorFlow/JAX
- you need lightweight statistical tools without a heavy framework dependency

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, math, data-science
- domain: machine-learning, deep-learning, data-science
- platform: python, cross-platform
- tags: probabilistic-programming, bayesian-inference, mcmc, variational-inference, tensorflow, jax, distributions, bijectors, statistics, gpu

## Member repositories
- tensorflow/probability (main) score 66

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:48.908481+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:20:56.461407+00:00, confidence not recorded.
  - readme: https://github.com/tensorflow/probability (fetched 2026-08-28T04:08:48.908481+00:00, sha 66006875f5bc)
  - homepage: https://www.tensorflow.org/probability/ (fetched 2026-08-29T09:08:34.626541+00:00, sha 3c2168cb2f55)
  - site_page: https://www.tensorflow.org/install (fetched 2026-08-29T09:08:34.670550+00:00, sha 584a762da891)
  - site_page: https://www.tensorflow.org/tfx/api_docs (fetched 2026-08-29T09:08:34.683599+00:00, sha 6977825696fe)
  - site_page: https://www.tensorflow.org/about (fetched 2026-08-29T09:08:34.687921+00:00, sha 817250744d91)
  - site_page: https://www.tensorflow.org/about/case-studies (fetched 2026-08-29T09:08:34.689811+00:00, sha e6701029eec0)
  - site_page: https://www.tensorflow.org/probability/api_docs/python/tfp (fetched 2026-08-29T09:08:34.691982+00:00, sha 6c4a2b884a7a)
  - site_page: https://www.tensorflow.org/about/bib (fetched 2026-08-29T09:08:34.693531+00:00, sha 5ca8943386b9)
  - site_page: https://www.tensorflow.org/community/contribute (fetched 2026-08-29T09:08:34.685949+00:00, sha 77d394fa76e7)
- Data as of 2026-08-30T08:39:29.467469+00:00.