Ei1jSSKrSSKJr SSKrSSKJr SSKJr SSKJr SSK J r S/r "SS5r g) N) deprecated)Tensor) constraints) lazy_property)_size Distributionc |^\rSrSrSrSrSrSr\S\ SS4Sj5r \ R"5\ R"5S4S \ RS \ RS \ S-SS4U4S jjjr S)S \4S jjr\S\ R4Sj5r\S\ R4Sj5r\S\\\R,44Sj5r\S\R,S-4Sj5r\S\4Sj5r\S\4Sj5r\S\4Sj5r\S\4Sj5r\ R"54S\S\4Sjjr\ R"54S\S\4Sjjr\ "S\!S9S\"S\4Sj5r#S\S\4Sjr$S\S\4Sjr%S\S\4Sjr&S*S \ S\4S!jjr'S\4S"jr(S\4S#jr)\ R"54S\S\ R4S$jjr*S\SS4S%jr+S)S&jr,S\4S'jr-S(r.U=r/$)+ra; Distribution is the abstract base class for probability distributions. Args: batch_shape (torch.Size): The shape over which parameters are batched. event_shape (torch.Size): The shape of a single sample (without batching). validate_args (bool, optional): Whether to validate arguments. Default: None. FTvaluereturnNc2US;a[eU[lg)a] Sets whether validation is enabled or disabled. The default behavior mimics Python's ``assert`` statement: validation is on by default, but is disabled if Python is run in optimized mode (via ``python -O``). Validation may be expensive, so you may want to disable it once a model is working. Args: value (bool): Whether to enable validation. )TFN) ValueErrorr_validate_args)r s ]/home/mande/repo/quber/.venv/lib/python3.13/site-packages/torch/distributions/distribution.pyset_default_validate_args&Distribution.set_default_validate_argss  % &+ # batch_shape event_shape validate_argsc>XlX lUbX0lUR(GaURnUR5HupV[R"U5(aM"XPR;a*[[[U5U5[5(aM[[X5nUR!U5n["R$"U5(aM['SUS[U5R(S[+UR,5S [/U5S [/U5S U3 5e [0T U]e5 g![a. 0n[ R "URS3S-S-SS9 GNAf=f) Nz$ does not define `arg_constraints`. zAPlease set `arg_constraints = {}` or initialize the distribution z2with `validate_args=False` to turn off validation. stacklevelzExpected parameter z ( of shape z) of distribution z to satisfy the constraint , but found invalid values: ) _batch_shape _event_shaperarg_constraintsNotImplementedErrorwarningswarn __class__itemsr is_dependent__dict__ isinstancegetattrtyperchecktorch _is_all_truer__name__tupleshapereprsuper__init__) selfrrrrparam constraintr validr#s rr2Distribution.__init__.sy ('  $"/      "&"6"6&5%:%:%<!++J77 -*DJ. 33,"((/))%00$-eW5 K001E%++&=" 3' "$ ~~&&JKYZJK !  s E4E87E8c[e)a Returns a new distribution instance (or populates an existing instance provided by a derived class) with batch dimensions expanded to `batch_shape`. This method calls :class:`~torch.Tensor.expand` on the distribution's parameters. As such, this does not allocate new memory for the expanded distribution instance. Additionally, this does not repeat any args checking or parameter broadcasting in `__init__.py`, when an instance is first created. Args: batch_shape (torch.Size): the desired expanded size. _instance: new instance provided by subclasses that need to override `.expand`. Returns: New distribution instance with batch dimensions expanded to `batch_size`. r )r3r _instances rexpandDistribution.expandVs &"!rcUR$)z6 Returns the shape over which parameters are batched. )rr3s rrDistribution.batch_shapek    rcUR$)z: Returns the shape of a single sample (without batching). )rr>s rrDistribution.event_shaperr@rc[e)z Returns a dictionary from argument names to :class:`~torch.distributions.constraints.Constraint` objects that should be satisfied by each argument of this distribution. Args that are not tensors need not appear in this dict. r9r>s rrDistribution.arg_constraintsy "!rc[e)zq Returns a :class:`~torch.distributions.constraints.Constraint` object representing this distribution's support. r9r>s rsupportDistribution.support "!rc[e)z' Returns the mean of the distribution. r9r>s rmeanDistribution.mean "!rc2[URS35e)z' Returns the mode of the distribution. z does not implement mode)r r#r>s rmodeDistribution.modes "T^^$44L"MNNrc[e)z+ Returns the variance of the distribution. r9r>s rvarianceDistribution.variancerMrc6URR5$)z5 Returns the standard deviation of the distribution. )rRsqrtr>s rstddevDistribution.stddevs }}!!##r sample_shapec[R"5 URU5sSSS5 $!,(df  g=f)z| Generates a sample_shape shaped sample or sample_shape shaped batch of samples if the distribution parameters are batched. N)r+no_gradrsampler3rXs rsampleDistribution.samples# ]]_<< -__s1 ?c[e)z Generates a sample_shape shaped reparameterized sample or sample_shape shaped batch of reparameterized samples if the distribution parameters are batched. r9r\s rr[Distribution.rsamplerIrz=`sample_n(n)` will be deprecated. Use `sample((n,))` instead.)categoryncNUR[R"U455$)zY Generates n samples or n batches of samples if the distribution parameters are batched. )r]r+Size)r3rbs rsample_nDistribution.sample_ns{{5::qd+,,rc[e)zk Returns the log of the probability density/mass function evaluated at `value`. Args: value (Tensor): r9r3r s rlog_probDistribution.log_probrErc[e)z_ Returns the cumulative density/mass function evaluated at `value`. Args: value (Tensor): r9rhs rcdfDistribution.cdfrErc[e)zg Returns the inverse cumulative density/mass function evaluated at `value`. Args: value (Tensor): r9rhs ricdfDistribution.icdfrErr;c[e)a Returns tensor containing all values supported by a discrete distribution. The result will enumerate over dimension 0, so the shape of the result will be `(cardinality,) + batch_shape + event_shape` (where `event_shape = ()` for univariate distributions). Note that this enumerates over all batched tensors in lock-step `[[0, 0], [1, 1], ...]`. With `expand=False`, enumeration happens along dim 0, but with the remaining batch dimensions being singleton dimensions, `[[0], [1], ..`. To iterate over the full Cartesian product use `itertools.product(m.enumerate_support())`. Args: expand (bool): whether to expand the support over the batch dims to match the distribution's `batch_shape`. Returns: Tensor iterating over dimension 0. r9)r3r;s renumerate_supportDistribution.enumerate_supports ,"!rc[e)zg Returns entropy of distribution, batched over batch_shape. Returns: Tensor of shape batch_shape. r9r>s rentropyDistribution.entropys "!rcJ[R"UR55$)zj Returns perplexity of distribution, batched over batch_shape. Returns: Tensor of shape batch_shape. )r+exprur>s r perplexityDistribution.perplexitysyy((rc[U[R5(d[R"U5n[R"XR-UR-5$)a@ Returns the size of the sample returned by the distribution, given a `sample_shape`. Note, that the batch and event shapes of a distribution instance are fixed at the time of construction. If this is empty, the returned shape is upcast to (1,). Args: sample_shape (torch.Size): the size of the sample to be drawn. )r'r+rdrrr\s r_extended_shapeDistribution._extended_shape sF, 33 ::l3Lzz,):)::T=N=NNOOrc [U[R5(d [S5e[ UR 55[ UR 5- nUR 5USUR :wa*[SUR 5SUR S35eUR 5nURUR -n[[U5[U55H,upVUS:wdM US:wdMXV:wdM[SUSUS35e URnUc [S 5eUR!U5n[R""U5(dR[S[%U5R&S[)UR*5S[-U5S[-U5SU3 5eg![a+ [R"URS3S -S -S S 9 gf=f)a Argument validation for distribution methods such as `log_prob`, `cdf` and `icdf`. The rightmost dimensions of a value to be scored via these methods must agree with the distribution's batch and event shapes. Args: value (Tensor): the tensor whose log probability is to be computed by the `log_prob` method. Raises ValueError: when the rightmost dimensions of `value` do not match the distribution's batch and event shapes. z/The value argument to log_prob must be a TensorNz5The right-most size of value must match event_shape: z vs .z9Value is not broadcastable with batch_shape+event_shape: z% does not define `support` to enable z;sample validation. Please initialize the distribution with z-`validate_args=False` to turn off validation.rrzsupport is unexpectedly NonezExpected value argument (rz) to be within the support (z) of the distribution r)r'r+rrlensizerrzipreversedrGr r!r"r#AssertionErrorr*r,r)r-r.r/r0) r3r event_dim_start actual_shapeexpected_shapeijrGr6s r_validate_sampleDistribution._validate_samples%..NO Oejjl+c$2C2C.DD ::<( )T->-> >G ~UYZ^ZkZkYllmn zz| **T->->>.0HIDAAv!q&QV OP\~]abpaqqrsJ llG ? !?@ @ e$!!%((K(()E%++4F3GH--1']O<''+Dzl3..3W 6 )#  MM>>""GHOPAB    s F662G+*G+cUcS[U5RUR:wa0[SURRSURS35eUcUR [U55$U$)Nz Subclass z of zR that defines a custom __init__ method must also define a custom .expand() method.)r)r2r r#r-__new__)r3clsr:s r_get_checked_instance"Distribution._get_checked_instanceLsp  d!4!4 !D%DNN334DG>> ,5+rs2( +3  MAMAr