'}h0ddlZddlmZddlmZmZmZmZmZmZm Z m Z m Z gdZ eddZ Gdd e e ZGd d eeZGd d eeZGddeeZGddeeZGddeeeZy)N)Tensor) IteratorIterableOptionalSequenceListTypeVarGenericSizedUnion) BatchSampler RandomSamplerSamplerSequentialSamplerSubsetRandomSamplerWeightedRandomSamplerT_coT) covariantc:eZdZdZddeeddfdZdeefdZ y)ra/Base 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 a :meth:`__len__` method that returns the length of the returned iterators. Args: data_source (Dataset): This argument is not used and will be removed in 2.2.0. You may still have custom implementation that utilizes it. 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`. N data_sourcereturnc4|ddl}|jdyy)Nrzz`data_source` argument is not used and will be removed in 2.2.0.You may still have custom implementation that utilizes it.)warningswarn)selfrrs W/var/www/html/test/engine/venv/lib/python3.12/site-packages/torch/utils/data/sampler.py__init__zSampler.__init__<s#  "  MMW X #ctN)NotImplementedErrorrs r__iter__zSampler.__iter__Cs!!rr ) __name__ __module__ __qualname____doc__rr rrrr#rrrrs2'RXHUOXtX"(4."rrcJeZdZUdZeed<deddfdZdeefdZ defdZ y)rz~Samples elements sequentially, always in the same order. Args: data_source (Dataset): dataset to sample from rrNc||_yr )r)rrs rrzSequentialSampler.__init__ks &rcPttt|jSr )iterrangelenrr"s rr#zSequentialSampler.__iter__nsE#d../011rc,t|jSr )r.rr"s r__len__zSequentialSampler.__len__qs4##$$r) r$r%r&r'r __annotations__rrintr#r0r(rrrrbs> 'E'd'2(3-2%%rrc ~eZdZUdZeed<eed< d dededeeddfdZ e defdZ de efd Z defd Zy) raSamples elements randomly. If without replacement, then sample from a shuffled dataset. If with replacement, then user can specify :attr:`num_samples` to draw. Args: data_source (Dataset): dataset to sample from replacement (bool): samples are drawn on-demand with replacement if ``True``, default=``False`` num_samples (int): number of samples to draw, default=`len(dataset)`. generator (Generator): Generator used in sampling. r replacementN num_samplesrc"||_||_||_||_t |jt st d|jt |jtr|jdkrtd|jy)N;replacement should be a boolean value, but got replacement=rDnum_samples should be a positive integer value, but got num_samples=) rr4 _num_samples generator isinstancebool TypeErrorr5r2 ValueError)rrr4r5r:s rrzRandomSampler.__init__s&&'"$**D1YZ^ZjZjYklm m$**C0D4D4D4Icdhdtdtcuvw w5Jrc\|jt|jS|jSr )r9r.rr"s rr5zRandomSampler.num_sampless-    $t''( (   rc#Kt|j}|jptt j dtj jj}t j}|j|n |j}|jrt|jdzD]?}t j|dtj |jEd{At j||jdzftj |jEd{yt|j|zD]/}t j ||jEd{1t j ||jd|j|zEd{y777H7 w)Nr(dtype )rC)highsizerBr:r:)r.rr:r2torchemptyint64random_item Generator manual_seedr4r-r5randinttolistrandperm)rnseedr:_s rr#zRandomSampler.__iter__sy   ! >> !u{{2U[[9AACHHJKD)I  ! !$ 'I   4++r12 n ==ae5;;Zcdkkmmm n}}!43C3Cb3H2JRWR]R]irszz| | |4++q01 K >>!yAHHJJJ K~~a9=DDFG\HXHX[\H\] ] ] n |K ]sJC=G"?GA G" G AG"G>G"G G"G"G" G"c|jSr r5r"s rr0zRandomSampler.__len__r)FNN)r$r%r&r'r r1r<rr2rpropertyr5rr#r0r(rrrrus ?D>B xE x x&sm xGK x!S!! ^(3-^$  rrcXeZdZUdZeeed<ddeeddfdZdeefdZ defdZ y) rzSamples elements randomly from a given list of indices, without replacement. Args: indices (sequence): a sequence of indices generator (Generator): Generator used in sampling. indicesNrc ||_||_yr )rYr:)rrYr:s rrzSubsetRandomSampler.__init__s "rc#Ktjt|j|jD]}|j|ywNrF)rGrPr.rYr:)ris rr#zSubsetRandomSampler.__iter__s;DLL 1T^^L "A,,q/ ! "sA Ac,t|jSr )r.rYr"s rr0zSubsetRandomSampler.__len__s4<<  rr ) r$r%r&r'rr2r1rrr#r0r(rrrrsFc]# #$#"(3-"!!rrc peZdZUdZeed<eed<eed< d dee dededdfdZ de efdZ defd Z y) raNSamples 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] weightsr5r4Nrct|trt|ts|dkrtd|t|tstd|t j |tj }t|jdk7r!tdt|j||_ ||_ ||_ ||_ y)Nrr8r7rAz=weights should be a 1d sequence but given weights have shape )r;r2r<r>rG as_tensordoubler.shapetupler`r5r4r:)rr`r5r4r:weights_tensors rrzWeightedRandomSampler.__init__s+s+z+t/Lq cdocpqr r+t,Z[fZghi i E ~## $ )3389M9M3N2OQR R& &&"rc#Ktj|j|j|j|j }t |jEd{y7wr\)rG multinomialr`r5r4r:r,rO)r rand_tensors rr#zWeightedRandomSampler.__iter__sJ'' d6F6FHXHXdhdrdrs  **,---sAA( A&!A(c|jSr rUr"s rr0zWeightedRandomSampler.__len__rVr)TN)r$r%r&r'rr1r2r<rfloatrrr#r0r(rrrrsb$O6:##c#"#>B#$.(3-.  rrcbeZdZdZdeeeeefdededdfdZ de e efdZ defd Z y) r aiWraps 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]] sampler batch_size drop_lastrNct|trt|ts|dkrtd|t|tstd|||_||_||_y)NrzBbatch_size should be a positive integer value, but got batch_size=z7drop_last should be a boolean value, but got drop_last=)r;r2r<r>rnrorp)rrnrorps rrzBatchSampler.__init__sg*c*jT.Jaablamno o)T*VW`Vabc c $"rc#K|jrGt|j} t|jDcgc] }t |}}|1dg|jz}d}|jD]2}|||<|dz }||jk(s|d}dg|jz}4|dkDr|d|yycc}w#t $rYywxYww)Nrrb)rpr,rnr-ronext StopIteration)r sampler_iterrSbatch idx_in_batchidxs rr#zBatchSampler.__iter__s >> -L9>t9OPAT,/PEPKC$//)EL|| 2&)l#! 4??2K#$LC$//1E  2aM\** Q$s?#C B<B7B<;C &C 7B<< CC CC c|jr"t|j|jzSt|j|jzdz |jzS)Nrb)rpr.rnror"s rr0zBatchSampler.__len__)sI >>t||$7 7 %7!;O Or)r$r%r&r'r rr2rr<rrrr#r0r(rrr r s`  #gclHSM&A B #PS #`d #im #+(49-+.PPrr )rGrtypingrrrrrr r r r __all__rrr2rrrrr r(rrr|s ___ v&2"gdm2"`% %&6 GCL6 r!'#,!,. GCL. b<P749%<Pr