o p i{E@slUddlmZddlZddlmZmZddlmZddlZ ddl mZ ddl Z ddl mZmZddlmZddlmZddlmZdd lmZerdd lmZdd lmZdd l mZmZdd lmZeee fZ!de"d<ee j#e j$e j%e j&fZ'de"d<ee!ee!ee!e j(e'efZ)de"d<eeeeeee j(ee j#e j$fefZ*de"d<Gdddej+Z,dS)) annotationsN)IterableSequence) TYPE_CHECKING) check_type convert_dtype)Variable) distribution)in_dynamic_mode)random)Union) TypeAlias)Tensordtype)NestedSequencer _NormalLocBase_NormalLocNDArray _NormalLoc _NormalScalecseZdZUdZded<ded<ded<ded< d'd(fdd Zed)ddZed)ddZgdfd*ddZ gfd+ddZ d)ddZ d,d d!Z d,d"d#Z d-d%d&ZZS).Normala The Normal distribution with location `loc` and `scale` parameters. Mathematical details If 'loc' is real number, the probability density function (pdf) is .. math:: pdf(x; \mu, \sigma) = \frac{1}{Z}e^{\frac {-0.5 (x - \mu)^2} {\sigma^2} } .. math:: Z = (2 \pi \sigma^2)^{0.5} If 'loc' is complex number, the probability density function (pdf) is .. math:: pdf(x; \mu, \sigma) = \frac{1}{Z}e^{\frac {-(x - \mu)^2} {\sigma^2} } .. math:: Z = \pi \sigma^2 In the above equations: * :math:`loc = \mu`: is the mean. * :math:`scale = \sigma`: is the std. * :math:`Z`: is the normalization constant. Args: loc(int|float|complex|list|tuple|numpy.ndarray|Tensor): The mean of normal distribution.The data type is float32, float64, complex64 and complex128. scale(int|float|list|tuple|numpy.ndarray|Tensor): The std of normal distribution.The data type is float32 and float64. name(str|None, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`. Examples: .. code-block:: python >>> import paddle >>> from paddle.distribution import Normal >>> # Define a single scalar Normal distribution. >>> dist = Normal(loc=0., scale=3.) >>> # Define a batch of two scalar valued Normals. >>> # The first has mean 1 and standard deviation 11, the second 2 and 22. >>> dist = Normal(loc=[1., 2.], scale=[11., 22.]) >>> # Get 3 samples, returning a 3 x 2 tensor. >>> dist.sample([3]) >>> # Define a batch of two scalar valued Normals. >>> # Both have mean 1, but different standard deviations. >>> dist = Normal(loc=1., scale=[11., 22.]) >>> # Complete example >>> value_tensor = paddle.to_tensor([0.8], dtype="float32") >>> normal_a = Normal([0.], [1.]) >>> normal_b = Normal([0.5], [2.]) >>> sample = normal_a.sample([2]) >>> # a random tensor created by normal distribution with shape: [2, 1] >>> entropy = normal_a.entropy() >>> print(entropy) Tensor(shape=[1], dtype=float32, place=Place(cpu), stop_gradient=True, [1.41893852]) >>> lp = normal_a.log_prob(value_tensor) >>> print(lp) Tensor(shape=[1], dtype=float32, place=Place(cpu), stop_gradient=True, [-1.23893857]) >>> p = normal_a.probs(value_tensor) >>> print(p) Tensor(shape=[1], dtype=float32, place=Place(cpu), stop_gradient=True, [0.28969154]) >>> kl = normal_a.kl_divergence(normal_b) >>> print(kl) Tensor(shape=[1], dtype=float32, place=Place(cpu), stop_gradient=True, [0.34939718]) rlocscalestrnamerNrr str | NonereturnNonec sBts&t|dttttjttj j t t fdt|dtttjttj j t t fdd|_ |dur/|nd|_d|_d|_t|trAt|}t|trJt|}t|t t frlt|}|jtjkra|d}|jtjkrl|d}t|t t fr{tj|tjd}t|tst|tjr|jtjtjfvs||r%|r%d|_t|trt|trd|_ t|tjr|jtjkrdnd }|jtjkrdnd }t|j|}t|j|}t|||_n t|trtj|jdd}tj|jdd}t|||_n||_t|tjr tj||jd|_nt|trtj|dd|_n||_t|jj|_nr|||r9||_||_t|j|_n^t|trHt|trHd|_ t|tjr\t |jd vr\|j|_nt|tjrot |jd vro|j|_|!||\|_|_|jt|jjkrtj"|j|jd|_tj"|j|jd|_t#$|jj%dS) NrrrFfloat32 complex64rTfloat64)rr )&r rintfloatcomplexnpZndarrayrpaddleZpirValuelisttupleall_arg_is_floatrr_complex_gaussian isinstancearrayr Zastype complex128rrZ_validate_argsZ is_complexZ to_tensorrealimagrrrrZ _to_tensorcastsuper__init__shape)selfrrrZ real_dtypeZ imag_dtyper.r/ __class__a/home/app/PaddleOCR-VL/.venv_paddleocr/lib/python3.10/site-packages/paddle/distribution/normal.pyr2s              zNormal.__init__cCs|jS)zWMean of normal distribution. Returns: Tensor: mean value. )rr4r7r7r8meansz Normal.meancCs |jdS)z_Variance of normal distribution. Returns: Tensor: variance value. )rpowr9r7r7r8variance s zNormal.variancerr3 Sequence[int]seedr!c CsXt|ts tdtst|dtdt|}t|j|jj }|j d}d|vrw||}t||}t |j|jd |d<t |d|j}t ||}t |} tj| |jr]dndd ||jd } | ||j} t j| |j|d } | S||}tj||jrdndd ||jd t j||jd |j} t j| |j|d } |jrt j| ||d S| S) a&Generate samples of the specified shape. Args: shape (Sequence[int], optional): Shape of the generated samples. seed (int): Python integer number. Returns: Tensor, A tensor with prepended dimensions shape.The data type is float32. %sample shape must be Iterable object.r?sampleZ_sampler?)r:Zstdr?rrr)r+r TypeErrorr rr!r'rrr3rr%itemfullrZreshaper Zgaussianr*addZzerosr)) r4r3r? batch_shaperZ output_shape fill_shapezero_tmpZzero_tmp_reshapeZzero_tmp_shapeZnormal_random_tmpoutputr7r7r8rAsL       z Normal.samplecCsHt|ts td|t|}tj|jrdnd|d}|j||j S)aGenerate reparameterized samples of the specified shape. Args: shape (Sequence[int], optional): Shape of the generated samples. Returns: Tensor: A tensor with prepended dimensions shape.The data type is float32. r@rDrC)r:r3) r+rrGZ _extend_shaper(r%normalr*rr)r4r3Zepsr7r7r8rsampleHs zNormal.rsamplecCs:|jd}t|j|jj}|jrUd|vr4t|}t|j|jd|d<|jj}t |d|}n t |d|jj}tj d|t t j dt |j||dSd|vr{t|}t|j|jd|d<|j|jj}t |d|}nt |d|j}tj d|dt d t j t |j||dS) aShannon entropy in nats. If non-complex, the entropy is .. math:: entropy(\sigma) = 0.5 \log (2 \pi e \sigma^2) If complex gaussian, the entropy is .. math:: entropy(\sigma) = \log (\pi e \sigma^2) + 1 In the above equation: * :math:`scale = \sigma`: is the std. Returns: Tensor, Shannon entropy of normal distribution.The data type is float32. Z_entropyrBrrCrE@rF?r;)rr'rrr3r*r%rHrrIrJmathlogpi)r4rrKrLZ fill_dtyperMr7r7r8entropy[s2 "zNormal.entropyvaluec Cs|jd}||j|}|j|j}t|j}|jr9tjd||j||j|d|t t j |dStjd||j||jd||t t dt j |dS)zLog probability density/mass function. Args: value (Tensor): The input tensor. Returns: Tensor: log probability.The data type is same with :attr:`value` . Z _log_probrQrF) r_check_values_dtype_in_probsrrr%rTr*subtractconjrSrUsqrt)r4rWrvarZ log_scaler7r7r8log_probs   zNormal.log_probcCs|jd}||j|}|j|j}|jr1tjtd||j||j|t j ||dStjtd||j||jd|t dt j |j|dS)zProbability density/mass function. Args: value (Tensor): The input tensor. Returns: Tensor, probability. The data type is same with :attr:`value` . Z_probsrXrFrQr;) rrYrrr*r%divideexpr[rSrUr\)r4rWrr]r7r7r8probss4   z Normal.probsothercCsts t|dtd|j|jkrtd|jd}|j|j}||}|j|j|j}|jr@||}||dt |S||}t j d|d|dt ||dS)aThe KL-divergence between two normal distributions. If non-complex, the KL-divergence is .. math:: KL\_divergence(\mu_0, \sigma_0; \mu_1, \sigma_1) = 0.5 (ratio^2 + (\frac{diff}{\sigma_1})^2 - 1 - 2 \ln {ratio}) If complex gaussian: .. math:: KL\_divergence(\mu_0, \sigma_0; \mu_1, \sigma_1) = ratio^2 + (\frac{diff}{\sigma_1})^2 - 1 - 2 \ln {ratio} .. math:: ratio = \frac{\sigma_0}{\sigma_1} .. math:: diff = \mu_1 - \mu_0 In the above equation: * :math:`loc = \mu_0`: is the mean of current Normal distribution. * :math:`scale = \sigma_0`: is the std of current Normal distribution. * :math:`loc = \mu_1`: is the mean of other Normal distribution. * :math:`scale = \sigma_1`: is the std of other Normal distribution. * :math:`ratio`: is the ratio of scales. * :math:`diff`: is the difference between means. Args: other (Normal): instance of Normal. Returns: Tensor, kl-divergence between two normal distributions.The data type is float32. rb kl_divergencezVThe kl divergence must be computed between two distributions in the same number field.Z_kl_divergencerErRrF) r rrr* ValueErrorrrrr[r%rTrJ)r4rbrZ var_ratiot1r7r7r8rcs&'    zNormal.kl_divergence)N)rrrrrrrr)rr)r3r>r?r!rr)r3r>rr)rWrrr)rbrrr)__name__ __module__ __qualname____doc____annotations__r2propertyr:r=rArPrVr^rarc __classcell__r7r7r5r8r:s$ Ns  5  5 #r)- __future__rrScollections.abcrrtypingrnumpyr$Z numpy.typingZnptr%Zpaddle.base.data_feederrrZpaddle.base.frameworkrZpaddle.distributionr Zpaddle.frameworkr Z paddle.tensorr r Ztyping_extensionsr rrZpaddle._typingrr"r#rrjrr rr-rZNDArrayrr Distributionrr7r7r7r8sL