inferpy.inference.variational.loss_functions package¶
Submodules¶
inferpy.inference.variational.loss_functions.elbo module¶
-
inferpy.inference.variational.loss_functions.elbo.ELBO(pvars, qvars, batch_weight=1, **kwargs)[source]¶ Compute the loss tensor from the expanded variables of p and q models. :param pvars: The dict with the expanded p random variables :type pvars: dict<inferpy.RandomVariable> :param qvars: The dict with the expanded q random variables :type qvars: dict<inferpy.RandomVariable> :param batch_weight: Weight to assign less importance to the energy, used when processing data in batches :type batch_weight: float
- Returns (tf.Tensor):
The generated loss tensor
Module contents¶
-
inferpy.inference.variational.loss_functions.ELBO(pvars, qvars, batch_weight=1, **kwargs)[source]¶ Compute the loss tensor from the expanded variables of p and q models. :param pvars: The dict with the expanded p random variables :type pvars: dict<inferpy.RandomVariable> :param qvars: The dict with the expanded q random variables :type qvars: dict<inferpy.RandomVariable> :param batch_weight: Weight to assign less importance to the energy, used when processing data in batches :type batch_weight: float
- Returns (tf.Tensor):
The generated loss tensor