fef0836b77dabd9a1816db90d0bb98ea2dd64ea0,test/examples/spectral_mixture_gp_regression_test.py,SpectralMixtureGPModel,__init__,#SpectralMixtureGPModel#,23

Before Change

        super(SpectralMixtureGPModel, self).__init__(GaussianLikelihood())
        self.mean_module = ConstantMean()
        self.covar_module = SpectralMixtureKernel()
        self.params = MLEParameterGroup(
            log_noise=BoundedParameter(torch.Tensor([-2]), -15, 15),
            log_mixture_weights=BoundedParameter(torch.zeros(3), -15, 15),
            log_mixture_means=BoundedParameter(torch.zeros(3), -15, 15),
            log_mixture_scales=BoundedParameter(torch.zeros(3), -15, 15)
        )

    def forward(self, x):
        mean_x = self.mean_module(x, constant=Variable(torch.Tensor([0])))
        covar_x = self.covar_module(x,

After Change

class SpectralMixtureGPModel(gpytorch.GPModel):
    def __init__(self):
        super(SpectralMixtureGPModel, self).__init__(GaussianLikelihood())
        self.mean_module = ConstantMean()
        self.covar_module = SpectralMixtureKernel()
        self.register_parameter("log_noise", nn.Parameter(torch.Tensor([-2])), bounds=(-5, 5))
        self.register_parameter("log_mixture_weights", nn.Parameter(torch.zeros(3)), bounds=(-5, 5))
        self.register_parameter("log_mixture_means", nn.Parameter(torch.zeros(3)), bounds=(-5, 5))
        self.register_parameter("log_mixture_scales", nn.Parameter(torch.zeros(3)), bounds=(-5, 5))

    def forward(self, x):
        mean_x = self.mean_module(x, constant=Variable(torch.Tensor([0])))
        covar_x = self.covar_module(x,

Instances