botorch 0.17.0

Bayesian Optimization in PyTorch

Stars: 3464, Watchers: 3464, Forks: 463, Open Issues: 96

The meta-pytorch/botorch repo was created 7 years ago and the last code push was 11 hours ago.
The project is very popular with an impressive 3464 github stars!

How to Install botorch

You can install botorch using pip

pip install botorch

or add it to a project with poetry

poetry add botorch

Package Details

Author

Meta Platforms, Inc.

License

None

Homepage

None

PyPi:

https://pypi.org/project/botorch/

Documentation:

https://botorch.org

GitHub Repo:

https://github.com/pytorch/botorch

Classifiers

• Scientific/Engineering

Related Packages

Errors

A list of common botorch errors.

Code Examples

Here are some botorch code examples and snippets.

GitHub Issues

The botorch package has 96 open issues on GitHub

• Replace deprecated torch.jit.trace with torch.compile
• Raise error for unsupported all_tasks in SaasFullyBayesianMultiTaskGP
• Enable eval of empty MTGP
• [Bug]: TruncatedMultivariateNormal does not respect nonzero batch size
• [Continued] Refactor botorch/sampling/pathwise and add support for product kernels
• Add new PFN as Ax model
• [Bug]: RuntimeError: Index out of bounds
• [Bug]: Docstring/typing for OptimizeAcqfInputs
• Slowdown in optimize_acqf with outcome constraints after upgrade to 0.16.1
• Adding assign to load_state_dict implementations
• Tutorial addition: Feasibility-driven trust Region Bayesian Optimization

See more issues on GitHub