# meta-pytorch/botorch

Bayesian optimization in PyTorch

Repository: https://github.com/meta-pytorch/botorch
Canonical: https://ross.abutalabs.com/products/botorch
Homepage: https://botorch.org/
Language: Jupyter Notebook
License: MIT
License Family: permissive
Last push: 2026-08-26T14:10:34+00:00

## Health v2 (maintenance only)
Score: 91/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 75, longevity 100
- inputs: {"age_days": 2956, "days_push": 7, "days_rel": 86, "gap_med": 72, "n_releases_24m": 12}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3589, forks 497 (observed 2026-08-28T04:08:11.184650+00:00)

## What it is
BoTorch is a Python library for Bayesian Optimization research built on top of PyTorch, part of the PyTorch ecosystem. It provides modular, extensible primitives for probabilistic models (via GPyTorch), Monte Carlo acquisition functions, and optimizers for sequential black-box optimization.

## Use cases
- optimize hyperparameters of machine learning models with Bayesian optimization
- implement custom acquisition functions for black-box optimization research
- fit Gaussian process surrogate models on GPU with PyTorch autograd
- run sequential optimization of costly-to-evaluate experiments
- build multi-task or multi-fidelity Bayesian optimization algorithms
- integrate Bayesian optimization into A/B testing or scientific experiments

## When to choose
- you are a researcher or sophisticated practitioner developing new Bayesian optimization algorithms
- you need low-level, modular control over models, acquisition functions, and optimizers
- you want PyTorch-native GPU support and autodifferentiation in your optimization stack
- you need Monte Carlo-based or batch acquisition functions without restrictive analytic assumptions

## When to avoid
- you just want a ready-made hyperparameter tuning or experiment platform - use Ax instead
- you are not familiar with Bayesian optimization or PyTorch
- you need a simple, high-level tuning library with minimal setup
- your problem is not suited to expensive black-box function optimization

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, math, sdk
- domain: machine-learning, data-science, artificial-intelligence
- platform: python, cross-platform
- tags: bayesian-optimization, pytorch, gaussian-processes, hyperparameter-optimization, black-box-optimization, acquisition-functions, gpytorch, research, algorithms, gpu

## Member repositories
- meta-pytorch/botorch (main) score 91

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:11.184650+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 2026-08-29T18:33:52.634148+00:00, confidence not recorded.
  - readme: https://github.com/meta-pytorch/botorch (fetched 2026-08-28T04:08:11.184650+00:00, sha 4a24c8a31f6d)
  - homepage: https://botorch.org/ (fetched 2026-08-29T09:27:26.052364+00:00, sha dd21f951292d)
  - site_page: https://botorch.org/docs/introduction (fetched 2026-08-29T09:27:26.061646+00:00, sha ecb9519b368e)
  - site_page: https://botorch.org/docs/tutorials (fetched 2026-08-29T09:27:26.063708+00:00, sha ca10f6bb1017)
  - site_page: https://botorch.org/docs/notebooks_community (fetched 2026-08-29T09:27:26.065545+00:00, sha c7101aef7e13)
  - site_page: https://botorch.org/docs/papers (fetched 2026-08-29T09:27:26.067159+00:00, sha 1c2c67fcded1)
  - site_page: https://botorch.org/docs/next/introduction (fetched 2026-08-29T09:27:26.068967+00:00, sha a6fda6979a52)
  - site_page: https://botorch.org/docs/v0.17.2/introduction (fetched 2026-08-29T09:27:26.070735+00:00, sha 3f55599cc7c0)
  - site_page: https://botorch.org/docs/v0.17.1/introduction (fetched 2026-08-29T09:27:26.072341+00:00, sha 3a704e857046)
  - site_page: https://botorch.org/docs/v0.17.0/introduction (fetched 2026-08-29T09:27:26.074073+00:00, sha 2cc0819cda98)
- Data as of 2026-08-30T08:39:29.467469+00:00.