+ É&j– ãóB€^RIt^RIHt^RIHtR.t!RR]4tR#)éN)ÚTensor)Ú DistributionÚExponentialFamilycóha€]tRt^ toRt]V3RlRl4tRt]V3RlRl4tRt Rt Vt R #) raš ExponentialFamily is the abstract base class for probability distributions belonging to an exponential family, whose probability mass/density function has the form is defined below .. math:: p_{F}(x; \theta) = \exp(\langle t(x), \theta\rangle - F(\theta) + k(x)) where :math:`\theta` denotes the natural parameters, :math:`t(x)` denotes the sufficient statistic, :math:`F(\theta)` is the log normalizer function for a given family and :math:`k(x)` is the carrier measure. Note: This class is an intermediary between the `Distribution` class and distributions which belong to an exponential family mainly to check the correctness of the `.entropy()` and analytic KL divergence methods. We use this class to compute the entropy and KL divergence using the AD framework and Bregman divergences (courtesy of: Frank Nielsen and Richard Nock, Entropies and Cross-entropies of Exponential Families). có4<€V^8„dQhRS[S[R3,/#)éÚreturn.)Útupler)ÚformatÚ __classdict__s"€Úv/Users/jameslopez/projects/CWCArchive/cwc-podcast/.venv/lib/python3.14/site-packages/torch/distributions/exp_family.pyÚ __annotate__ÚExponentialFamily.__annotate__!sø€÷"ñ"¡¡v¨s {Õ!3ñ"óc ó€\h)z^ Abstract method for natural parameters. Returns a tuple of Tensors based on the distribution ©ÚNotImplementedError©Úselfs&r Ú_natural_paramsÚ!ExponentialFamily._natural_params ó €ô "Ð!rc ó€\h)zk Abstract method for log normalizer function. Returns a log normalizer based on the distribution and input r)rÚnatural_paramss&*r Ú_log_normalizerÚ!ExponentialFamily._log_normalizer(s €ô "Ð!rcó <€V^8„dQhRS[/#)rr )Úfloat)r r s"€r rr0sø€÷"ñ"¡uñ"rc ó€\h)zX Abstract method for expected carrier measure, which is required for computing entropy. rrs&r Ú_mean_carrier_measureÚ'ExponentialFamily._mean_carrier_measure/rrc óÊ€VP)pVPUu.uF q"P4P4NK" ppVP!V!p\ P PVP4VRR7pW, p\W54FCwrgWV,PVPR,4PR4,pKE V#uupi)zO Method to compute the entropy using Bregman divergence of the log normalizer. T)Ú create_graphéÿÿÿÿ)r$) r rÚdetachÚrequires_grad_rÚtorchÚautogradÚgradÚsumÚzipÚreshapeÚ _batch_shape)rÚresultÚpÚnparamsÚ lg_normalÚ gradientsÚnpÚgs& r ÚentropyÚExponentialFamily.entropy7s½€ð#'×"<Ñ"<Ð!<ˆØ8<×8LÒ8LÓMÑ8L°1—8‘8“:×,Ñ,Ö.Ñ8LˆÐMØ×(Ò(¨'Ñ2ˆ Ü—N‘N×'Ñ'¨ ¯ © «¸ÈtÐ'ÓTˆ ØՈܘÖ,‰EˆBØ ˜A•v×&Ñ& t×'8Ñ'8¸5Õ'@ÓA×EÑEÀbÓIÕ IŠFñ-àˆ ùò Nsœ&C ©N) Ú__name__Ú __module__Ú __qualname__Ú__firstlineno__Ú__doc__Úpropertyrrr r5Ú__static_attributes__Ú__classdictcell__)r s@r rr sAø‡€ñð(÷"óð"ò"ð÷"óð"÷ ð r)r'rÚ torch.distributions.distributionrÚ__all__rr7rr ÚrBs&ðó ÝÝ9ð Ð €ô7˜ ö7r