9c526695805c9639896b31364958d0e77bdeba62,gpytorch/likelihoods/multitask_gaussian_likelihood.py,MultitaskGaussianLikelihood,__init__,#MultitaskGaussianLikelihood#Any#Any#Any#Any#Any#,27

Before Change

        
        super(MultitaskGaussianLikelihood, self).__init__(batch_size=batch_size, log_noise_prior=log_noise_prior)

        if rank == 0:
            self.register_parameter(
                name="log_task_noises",
                parameter=torch.nn.Parameter(torch.zeros(batch_size, num_tasks)),
                prior=task_prior,
            )
        else:
            self.register_parameter(
                name="task_noise_covar_factor", parameter=torch.nn.Parameter(torch.randn(batch_size, num_tasks, rank))
            )
            if task_prior is not None:
                self.register_derived_prior(
                    name="MultitaskErrorCovariancePrior",
                    prior=task_prior,
                    parameter_names=("task_noise_covar_factor", "log_noise"),
                    transform=_eval_covar_matrix,
                )
        self.num_tasks = num_tasks

    def forward(self, input):
        

After Change

            then a diagonal covariance matrix is fit.

            task_prior (:obj:`gpytorch.priors.Prior`): Prior to use over the task noise covariance matrix if
            `rank` > 0, or a prior over the log of just the diagonal elements, if `rank` == 0.

        
        super(MultitaskGaussianLikelihood, self).__init__(batch_size=batch_size, log_noise_prior=log_noise_prior)

        if rank == 0:
            self.register_parameter(
                name="log_task_noises",
                parameter=torch.nn.Parameter(torch.zeros(batch_size, num_tasks)),
                prior=task_prior,
            )
        else:
            self.register_parameter(
                name="task_noise_covar_factor", parameter=torch.nn.Parameter(torch.randn(batch_size, num_tasks, rank))
            )
            if task_prior is not None:
                self.register_derived_prior(
                    name="MultitaskErrorCovariancePrior",
                    prior=task_prior,
                    parameter_names=("task_noise_covar_factor", "log_noise"),
                    transform=self._eval_covar_matrix,

Instances