+ &j1@^RIt^RIHtHtHtHt^RIHtHt^RI t .ROt ]!R R R 7t !R R]] ,4t !RR] ] ,4t!RR] ] ,4t!RR] ] ,4t!RR ] ] ,4t!RR] ]] ,,4tR#)N)IterableIteratorSequenceSized)GenericTypeVar BatchSampler RandomSamplerSamplerSequentialSamplerSubsetRandomSamplerWeightedRandomSampler_T_coT) covariantc6a]tRt^toRtV3RlRltRtVtR#)r aBase class for all Samplers. Every Sampler subclass has to provide an :meth:`__iter__` method, providing a way to iterate over indices or lists of indices (batches) of dataset elements, and may provide a :meth:`__len__` method that returns the length of the returned iterators. Example: >>> # xdoctest: +SKIP >>> class AccedingSequenceLengthSampler(Sampler[int]): >>> def __init__(self, data: List[str]) -> None: >>> self.data = data >>> >>> def __len__(self) -> int: >>> return len(self.data) >>> >>> def __iter__(self) -> Iterator[int]: >>> sizes = torch.tensor([len(x) for x in self.data]) >>> yield from torch.argsort(sizes).tolist() >>> >>> class AccedingSequenceLengthBatchSampler(Sampler[List[int]]): >>> def __init__(self, data: List[str], batch_size: int) -> None: >>> self.data = data >>> self.batch_size = batch_size >>> >>> def __len__(self) -> int: >>> return (len(self.data) + self.batch_size - 1) // self.batch_size >>> >>> def __iter__(self) -> Iterator[List[int]]: >>> sizes = torch.tensor([len(x) for x in self.data]) >>> for batch in torch.chunk(torch.argsort(sizes), len(self)): >>> yield batch.tolist() .. note:: The :meth:`__len__` method isn't strictly required by :class:`~torch.utils.data.DataLoader`, but is expected in any calculation involving the length of a :class:`~torch.utils.data.DataLoader`. c0<V^8dQhRS[S[,/#return)rr)format __classdict__s"p/Users/jameslopez/projects/CWCArchive/cwc-podcast/.venv/lib/python3.14/site-packages/torch/utils/data/sampler.py __annotate__Sampler.__annotate__Bs""(5/"c\hN)NotImplementedErrorselfs&r__iter__Sampler.__iter__Bs!!rN)__name__ __module__ __qualname____firstlineno____doc__r!__static_attributes____classdictcell__rs@rr r s#J""rcfa]tRt^atoRtV3RlRltV3RlRltV3RlRltV3RltR t Vt R #) r zSamples elements sequentially, always in the same order. Args: data_source (Sized): data source to sample from. Must implement __len__. c$<V^8dQhRS[RR/#)r data_sourcerNr)rrs"rrSequentialSampler.__annotate__js''E'd'rcWnR#rr.)r r.s&&r__init__SequentialSampler.__init__js&rc0<V^8dQhRS[S[,/#rrint)rrs"rrr0ms22(3-2rcP\\\VP444#r)iterrangelenr.rs&rr!SequentialSampler.__iter__msE#d../011rc <V^8dQhRS[/#rr7)rrs"rrr0ps%%%rc,\VP4#r)r;r.rs&r__len__SequentialSampler.__len__ps4##$$rc&<V^8dQh/S[;R&#)rr.r/)rrs"rrr0asrr2N r$r%r&r'r(r3r!r@__annotate_func__r)r*r+s@rr r as/''22%%rca]tRt^ttoRtR V3RlRllt]V3RlRl4tV3RlRltV3R lR lt V3R lt R t Vt R#)r aSamples elements randomly. If without replacement, then sample from a shuffled dataset. If with replacement, then user can specify :attr:`num_samples` to draw. Args: data_source (Sized): data source to sample from. Must implement __len__. replacement (bool): samples are drawn on-demand with replacement if ``True``, default=``False`` num_samples (int): number of samples to draw, default=`len(dataset)`. generator (Generator): Generator used in sampling. Nc><V^8dQhRS[RS[RS[R,RR/#)rr. replacement num_samplesNr)rboolr7)rrs"rrRandomSampler.__annotate__s74Z   rc8WnW nW0nW@n\ VP\ 4'g\ RVP 24h\ VP\4'dVP^8:d\RVP 24hR#);replacement should be a boolean value, but got replacement=Dnum_samples should be a positive integer value, but got num_samples=N) r.rG _num_samples generator isinstancerI TypeErrorrHr7 ValueError)r r.rGrHrOs&&&&&rr3RandomSampler.__init__s'&'"$**D11MdN^N^M_` $**C00D4D4D4IVW[WgWgVhi 5Jrc <V^8dQhRS[/#rr>)rrs"rrrJs!!S!rc`VPf\VP4#VP#r)rNr;r.rs&rrHRandomSampler.num_sampless-    $t''( (   rc0<V^8dQhRS[S[,/#rr6)rrs"rrrJs(3-rc# "\VP4pVPfs\\P !R\P R7P4P44p\P!4pVPV4M VPpVP'd\VP^ ,4FAp\P!VR\P VR7P4RjxL KC \P!VVP^ ,3\P VR7P4RjxL R#\VPV,4F0p\P !WR7P4RjxL K2 \P !WR7P4RVPV,RjxL R#LLLOL 5i)Ndtype)highsizerZrOrOr#) )r;r.rOr7torchemptyint64random_item Generator manual_seedrGr:rHrandinttolistrandperm)r nseedrO_s& rr!RandomSampler.__iter__s|   ! >> !u{{2U[[9AACHHJKD)I  ! !$ 'I    4++r12 ==ekkY&(3}}&&+-kk#  fh   4++q01 >>!AHHJJJ2~~a=DDF&$""Q&   K sRB3H6AHG<AH"G>#A H0H1AH5H6H>HHHc <V^8dQhRS[/#rr>)rrs"rrrJ   rcVP#rrHrs&rr@RandomSampler.__len__rc2<V^8dQh/S[;R&S[;R&#)rr.rG)rrI)rrs"rrrJtsr)rNr.rOrG)FNN) r$r%r&r'r(r3propertyrHr!r@rDr)r*r+s@rr r tsD ,!! 6  Orcja]tRt^toRtR V3RlRlltV3RlRltV3RlRltV3R ltR t Vt R#) r zSamples elements randomly from a given list of indices, without replacement. Args: indices (sequence): a sequence of indices generator (Generator): Generator used in sampling. Nc4<V^8dQhRS[S[,RR/#)rindicesrNrr7)rrs"rr SubsetRandomSampler.__annotate__s## #$#rcWnW nR#r)rwrO)r rwrOs&&&rr3SubsetRandomSampler.__init__s  "rc0<V^8dQhRS[S[,/#rr6)rrs"rrrys""(3-"rc#"\P!\VP4VPR7P 4FpVPV,xK R#5ir]N)r_rhr;rwrOrg)r is& rr!SubsetRandomSampler.__iter__s?DLL 1T^^LSSUA,,q/ !VsA"A$c <V^8dQhRS[/#rr>)rrs"rrrys!!!rc,\VP4#r)r;rwrs&rr@SubsetRandomSampler.__len__s4<<  rc6<V^8dQh/S[S[,;R&#)rrwrx)rrs"rrrysc]r)rOrwrrCr+s@rr r s/##""!!%rcja]tRt^toRtR V3RlRlltV3RlRltV3RlRltV3R ltR t Vt R#) ra~Samples elements from ``[0,..,len(weights)-1]`` with given probabilities (weights). Args: weights (sequence) : a sequence of weights, not necessary summing up to one num_samples (int): number of samples to draw replacement (bool): if ``True``, samples are drawn with replacement. If not, they are drawn without replacement, which means that when a sample index is drawn for a row, it cannot be drawn again for that row. generator (Generator): Generator used in sampling. Example: >>> # xdoctest: +IGNORE_WANT("non-deterministic") >>> list( ... WeightedRandomSampler( ... [0.1, 0.9, 0.4, 0.7, 3.0, 0.6], 5, replacement=True ... ) ... ) [4, 4, 1, 4, 5] >>> list( ... WeightedRandomSampler( ... [0.9, 0.4, 0.05, 0.2, 0.3, 0.1], 5, replacement=False ... ) ... ) [0, 1, 4, 3, 2] Nc@<V^8dQhRS[S[,RS[RS[RR/#)rweightsrHrGrN)rfloatr7rI)rrs"rr"WeightedRandomSampler.__annotate__s7##%## #  #rc\V\4'd\V\4'gV^8:d\RV 24h\V\4'g\RV 24h\P !V\P R7p\VP4^8wd"\R\VP4 24hWPn W n W0n W@n R#)rrMrLrYz=weights should be a 1d sequence but given weights have shape N)rPr7rIrRr_ as_tensordoubler;shapetuplerrHrGrO)r rrHrGrOweights_tensors&&&&& rr3WeightedRandomSampler.__init__s;,,+t,,aVWbVcd +t,,Mk][  E ~## $ )&&+N,@,@&A%BD  & &&"rc0<V^8dQhRS[S[,/#rr6)rrs"rrrs..(3-.rc#"\P!VPVPVPVP R7p\ VP44RjxL R#L5ir~)r_ multinomialrrHrGrOr9rg)r rand_tensors& rr!WeightedRandomSampler.__iter__sK'' LL$**D,<,<  **,---sAA*!A("A*c <V^8dQhRS[/#rr>)rrs"rrrrnrcVP#rrprs&rr@WeightedRandomSampler.__len__rrrcR<V^8dQh/S[P;R&S[;R&S[;R&#)rrrHrG)r_Tensorr7rI)rrs"rrrs06\\789:;r)rOrHrGr)TNrCr+s@rrrs1<##@..   KrcZa]tRtRtoRtV3RlRltV3RlRltV3RlRltR tVt R #) r iaWraps another sampler to yield a mini-batch of indices. Args: sampler (Sampler or Iterable): Base sampler. Can be any iterable object batch_size (int): Size of mini-batch. drop_last (bool): If ``True``, the sampler will drop the last batch if its size would be less than ``batch_size`` Example: >>> list( ... BatchSampler( ... SequentialSampler(range(10)), batch_size=3, drop_last=False ... ) ... ) [[0, 1, 2], [3, 4, 5], [6, 7, 8], [9]] >>> list( ... BatchSampler(SequentialSampler(range(10)), batch_size=3, drop_last=True) ... ) [[0, 1, 2], [3, 4, 5], [6, 7, 8]] c`<V^8dQhRS[S[,S[S[,,RS[RS[RR/#)rsampler batch_size drop_lastrN)r r7rrI)rrs"rrBatchSampler.__annotate__4s@## -## #  #rc\V\4'd\V\4'gV^8:d\RV 24h\V\4'g\RV 24hWnW nW0nR#)rzBbatch_size should be a positive integer value, but got batch_size=z7drop_last should be a boolean value, but got drop_last=N)rPr7rIrRrrr)r rrrs&&&&rr3BatchSampler.__init__4su:s++*d++QTU_T`a )T**I)U  $"rc@<V^8dQhRS[S[S[,,/#r)rlistr7)rrs"rrrMs K K(49- Krc#\"\VP4pVP'd/V.VP,p\ VRR/F p.VOxK R#.\ P !WP4OpV'd)Vx.\ P !WP4OpK0R#5i)strictFN)r9rrrzip itertoolsislice)r sampler_iterargsbatch_droplastbatchs& rr!BatchSampler.__iter__MsDLL) >>> >DOO3D"%t":E":'''#;Gi&&|__EFE J)**<IJs A?B,*B,c <V^8dQhRS[/#rr>)rrs"rrrZsPPPrcVP'd'\VP4VP,#\VP4VP,^, VP,#))rr;rrrs&rr@BatchSampler.__len__ZsK >>>t||$7 7 %7!;O Or)rrrN) r$r%r&r'r(r3r!r@r)r*r+s@rr r s,*##2 K KPPr)r r r r r r)rcollections.abcrrrrtypingrrr___all__rr r7r r r rrr r#rrrs??#   4('"gen'"J% %&H GCLH V!'#,!,F GCLF RDP749%DPr