botorch.fit¶

Utilities for model fitting.

botorch.fit.fit_gpytorch_model(mll, optimizer=<function fit_gpytorch_scipy>, **kwargs)[source]¶

Fit hyperparameters of a gpytorch model.

Optimizer functions are in botorch.optim.fit.

Parameters:	mll (`MarginalLogLikelihood`) – MarginalLogLikelihood to be maximized. optimizer (`Callable`) – The optimizer function. kwargs (`Any`) – Arguments passed along to the optimizer function.
Return type:	`MarginalLogLikelihood`
Returns:	MarginalLogLikelihood with optimized parameters.

Example

>>> gp = SingleTaskGP(train_X, train_Y)
>>> mll = ExactMarginalLogLikelihood(gp.likelihood, gp)
>>> fit_gpytorch_model(mll)

BoTorch