157041a86ecda19d43967fc3028d7b48fe17b243,gpytorch/variational/variational_strategy.py,VariationalStrategy,prior_distribution,#VariationalStrategy#,63

Before Change

        this is done simply by calling the user defined GP prior on the inducing point data directly.
        
        out = self.model.forward(self.inducing_points)
        res = MultivariateNormal(
            out.mean, out.lazy_covariance_matrix.add_jitter()
        )
        return res

    def kl_divergence(self):
        variational_dist_u = self.variational_distribution.variational_distribution

After Change

    @property
    @cached(name="prior_distribution_memo")
    def prior_distribution(self):
        zeros = torch.zeros_like(self.inducing_points[..., 0])
        ones = torch.ones_like(zeros)
        res = MultivariateNormal(zeros, DiagLazyTensor(ones))
        return res

    def forward(self, x):

Instances