U MhJ@sddlZddlZddlZddlZddlmZmZmZmZm Z m Z m Z m Z m Z mZddlmZddlmZmZddlmZmZddd d d d d dgZe dddZe dZeeefZe edfZe deeZGdddeeZGdddeeeeZGdd d ee edfZ Gdd d eeZ!Gdd d eeZ"Gdd d eZ#Gdd d eeZ$efeee ee%e&fe ee e$edddZ'dS)N) castDictGenericIterableListOptionalSequenceTupleTypeVarUnion) deprecated)default_generatorrandperm) GeneratorTensorDatasetIterableDataset TensorDataset StackDataset ConcatDataset ChainDatasetSubset random_splitT_coT) covariantT.T_stackc@s.eZdZdZedddZddddd Zd S) raAn abstract class representing a :class:`Dataset`. All datasets that represent a map from keys to data samples should subclass it. All subclasses should overwrite :meth:`__getitem__`, supporting fetching a data sample for a given key. Subclasses could also optionally overwrite :meth:`__len__`, which is expected to return the size of the dataset by many :class:`~torch.utils.data.Sampler` implementations and the default options of :class:`~torch.utils.data.DataLoader`. Subclasses could also optionally implement :meth:`__getitems__`, for speedup batched samples loading. This method accepts list of indices of samples of batch and returns list of samples. .. note:: :class:`~torch.utils.data.DataLoader` by default constructs an index sampler that yields integral indices. To make it work with a map-style dataset with non-integral indices/keys, a custom sampler must be provided. )returncCs tddS)Nz3Subclasses of Dataset should implement __getitem__.)NotImplementedErrorselfindexr#J/var/www/html/venv/lib/python3.8/site-packages/torch/utils/data/dataset.py __getitem__>szDataset.__getitem__z Dataset[T_co]zConcatDataset[T_co])otherrcCs t||gSN)rr!r&r#r#r$__add__EszDataset.__add__N)__name__ __module__ __qualname____doc__rr%r)r#r#r#r$r+sc@s"eZdZdZeedddZdS)raHAn iterable Dataset. All datasets that represent an iterable of data samples should subclass it. Such form of datasets is particularly useful when data come from a stream. All subclasses should overwrite :meth:`__iter__`, which would return an iterator of samples in this dataset. When a subclass is used with :class:`~torch.utils.data.DataLoader`, each item in the dataset will be yielded from the :class:`~torch.utils.data.DataLoader` iterator. When :attr:`num_workers > 0`, each worker process will have a different copy of the dataset object, so it is often desired to configure each copy independently to avoid having duplicate data returned from the workers. :func:`~torch.utils.data.get_worker_info`, when called in a worker process, returns information about the worker. It can be used in either the dataset's :meth:`__iter__` method or the :class:`~torch.utils.data.DataLoader` 's :attr:`worker_init_fn` option to modify each copy's behavior. Example 1: splitting workload across all workers in :meth:`__iter__`:: >>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_DATALOADER) >>> # xdoctest: +SKIP("Fails on MacOS12") >>> class MyIterableDataset(torch.utils.data.IterableDataset): ... def __init__(self, start, end): ... super(MyIterableDataset).__init__() ... assert end > start, "this example code only works with end >= start" ... self.start = start ... self.end = end ... ... def __iter__(self): ... worker_info = torch.utils.data.get_worker_info() ... if worker_info is None: # single-process data loading, return the full iterator ... iter_start = self.start ... iter_end = self.end ... else: # in a worker process ... # split workload ... per_worker = int(math.ceil((self.end - self.start) / float(worker_info.num_workers))) ... worker_id = worker_info.id ... iter_start = self.start + worker_id * per_worker ... iter_end = min(iter_start + per_worker, self.end) ... return iter(range(iter_start, iter_end)) ... >>> # should give same set of data as range(3, 7), i.e., [3, 4, 5, 6]. >>> ds = MyIterableDataset(start=3, end=7) >>> # Single-process loading >>> print(list(torch.utils.data.DataLoader(ds, num_workers=0))) [tensor([3]), tensor([4]), tensor([5]), tensor([6])] >>> # xdoctest: +REQUIRES(POSIX) >>> # Mult-process loading with two worker processes >>> # Worker 0 fetched [3, 4]. Worker 1 fetched [5, 6]. >>> # xdoctest: +IGNORE_WANT("non deterministic") >>> print(list(torch.utils.data.DataLoader(ds, num_workers=2))) [tensor([3]), tensor([5]), tensor([4]), tensor([6])] >>> # With even more workers >>> # xdoctest: +IGNORE_WANT("non deterministic") >>> print(list(torch.utils.data.DataLoader(ds, num_workers=12))) [tensor([3]), tensor([5]), tensor([4]), tensor([6])] Example 2: splitting workload across all workers using :attr:`worker_init_fn`:: >>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_DATALOADER) >>> class MyIterableDataset(torch.utils.data.IterableDataset): ... def __init__(self, start, end): ... super(MyIterableDataset).__init__() ... assert end > start, "this example code only works with end >= start" ... self.start = start ... self.end = end ... ... def __iter__(self): ... return iter(range(self.start, self.end)) ... >>> # should give same set of data as range(3, 7), i.e., [3, 4, 5, 6]. >>> ds = MyIterableDataset(start=3, end=7) >>> # Single-process loading >>> print(list(torch.utils.data.DataLoader(ds, num_workers=0))) [3, 4, 5, 6] >>> >>> # Directly doing multi-process loading yields duplicate data >>> print(list(torch.utils.data.DataLoader(ds, num_workers=2))) [3, 3, 4, 4, 5, 5, 6, 6] >>> # Define a `worker_init_fn` that configures each dataset copy differently >>> def worker_init_fn(worker_id): ... worker_info = torch.utils.data.get_worker_info() ... dataset = worker_info.dataset # the dataset copy in this worker process ... overall_start = dataset.start ... overall_end = dataset.end ... # configure the dataset to only process the split workload ... per_worker = int(math.ceil((overall_end - overall_start) / float(worker_info.num_workers))) ... worker_id = worker_info.id ... dataset.start = overall_start + worker_id * per_worker ... dataset.end = min(dataset.start + per_worker, overall_end) ... >>> # Mult-process loading with the custom `worker_init_fn` >>> # Worker 0 fetched [3, 4]. Worker 1 fetched [5, 6]. >>> print(list(torch.utils.data.DataLoader(ds, num_workers=2, worker_init_fn=worker_init_fn))) [3, 5, 4, 6] >>> # With even more workers >>> print(list(torch.utils.data.DataLoader(ds, num_workers=12, worker_init_fn=worker_init_fn))) [3, 4, 5, 6] )r&cCs t||gSr')rr(r#r#r$r)szIterableDataset.__add__N)r*r+r,r-rrr)r#r#r#r$rMslc@sBeZdZUdZeedfed<eddddZdd Zd d Z dS) rzDataset wrapping tensors. Each sample will be retrieved by indexing tensors along the first dimension. Args: *tensors (Tensor): tensors that have the same size of the first dimension. .tensorsN)r.rcs(tfddDstd|_dS)Nc3s&|]}dd|dkVqdS)rN)size.0Ztensorr.r#r$ sz)TensorDataset.__init__..zSize mismatch between tensors)allAssertionErrorr.)r!r.r#r2r$__init__s   zTensorDataset.__init__cstfdd|jDS)Nc3s|]}|VqdSr'r#r0r"r#r$r3sz,TensorDataset.__getitem__..)tupler.r r#r7r$r%szTensorDataset.__getitem__cCs|jddSNr)r.r/r!r#r#r$__len__szTensorDataset.__len__) r*r+r,r-r r__annotations__r6r%r;r#r#r#r$rs c@sZeZdZUdZeeefed<ee ee ddddZ ddZ e d d d Z d d ZdS)raDataset as a stacking of multiple datasets. This class is useful to assemble different parts of complex input data, given as datasets. Example: >>> # xdoctest: +SKIP >>> images = ImageDataset() >>> texts = TextDataset() >>> tuple_stack = StackDataset(images, texts) >>> tuple_stack[0] == (images[0], texts[0]) >>> dict_stack = StackDataset(image=images, text=texts) >>> dict_stack[0] == {'image': images[0], 'text': texts[0]} Args: *args (Dataset): Datasets for stacking returned as tuple. **kwargs (Dataset): Datasets for stacking returned as dict. datasetsN)argskwargsrcs|rD|rtdt|d_tfdd|Dr.zSize mismatch between datasetsc3s|]}jt|kVqdSr'r@rCr:r#r$r3sz%At least one dataset should be passed) ValueErrorrBrAanyr=listvalues)r!r>r?tmpr#r:r$r6s  zStackDataset.__init__cs<t|jtr$fdd|jDStfdd|jDS)Ncsi|]\}}||qSr#r#)r1krDr7r#r$ sz,StackDataset.__getitem__..c3s|]}|VqdSr'r#rCr7r#r$r3sz+StackDataset.__getitem__..) isinstancer=dictitemsr8r r#r7r$r%s zStackDataset.__getitem__indicesc Csrt|jtrdd|D}|jD]\}}tt|ddr||}t|t|krrtdt|dt|t ||D]\}}|||<q|q$t ||D]\}}||||<qq$|Sdd|D} |jD]}tt|ddr:||}t|t|krtdt|dt|t || D]\}} | |q"qt || D]\}} | ||qDqdd| D} | S)NcSsg|]}iqSr#r#r1_r#r#r$ sz-StackDataset.__getitems__.. __getitems__z0Nested dataset's output size mismatch. Expected z, got cSsg|]}gqSr#r#rQr#r#r$rSscSsg|] }t|qSr#)r8)r1sampler#r#r$rS)s) rLr=rMrNcallablegetattrrTrBrEzipappend) r!rPZ dict_batchrJrDrNdataZd_sampleidxZ list_batchZt_sampleZ tuple_batchr#r#r$rTs8     zStackDataset.__getitems__cCs|jSr')rAr:r#r#r$r;,szStackDataset.__len__)r*r+r,r-r r8rMr<rrr6r%rGrTr;r#r#r#r$rs %cs~eZdZUdZeeeed<eeed<e ddZ e eddfdd Z d d Z d d ZeededddZZS)rzDataset as a concatenation of multiple datasets. This class is useful to assemble different existing datasets. Args: datasets (sequence): List of datasets to be concatenated r=cumulative_sizescCs6gd}}|D]"}t|}|||||7}q|Sr9)rBrY)sequencerselr#r#r$cumsum<s   zConcatDataset.cumsumNr=rcsZtt||_t|jdks*td|jD]}t|tr0tdq0||j|_ dS)Nrz(datasets should not be an empty iterablez.ConcatDataset does not support IterableDataset) superr6rGr=rBr5rLrrbr\)r!r=d __class__r#r$r6Es   zConcatDataset.__init__cCs |jdS)Nr\r:r#r#r$r;OszConcatDataset.__len__cCsf|dkr*| t|krtdt||}t|j|}|dkrF|}n||j|d}|j||S)Nrz8absolute value of index should not exceed dataset length)rBrEbisect bisect_rightr\r=)r!r[Z dataset_idxZ sample_idxr#r#r$r%Rs zConcatDataset.__getitem__z>`cummulative_sizes` attribute is renamed to `cumulative_sizes`)categorycCs|jSr'rir:r#r#r$cummulative_sizes`szConcatDataset.cummulative_sizes)r*r+r,r-rrrr<int staticmethodrbrr6r;r%propertyr FutureWarningrn __classcell__r#r#rfr$r0s    cs<eZdZdZeeddfdd ZddZdd ZZ S) ra_Dataset for chaining multiple :class:`IterableDataset` s. This class is useful to assemble different existing dataset streams. The chaining operation is done on-the-fly, so concatenating large-scale datasets with this class will be efficient. Args: datasets (iterable of IterableDataset): datasets to be chained together Nrccst||_dSr')rdr6r=)r!r=rfr#r$r6ts zChainDataset.__init__ccs,|jD] }t|tstd|EdHqdS)N*ChainDataset only supports IterableDataset)r=rLrr5)r!rer#r#r$__iter__xs zChainDataset.__iter__cCs2d}|jD]"}t|ts td|t|7}q |S)Nrrt)r=rLrr5rB)r!totalrer#r#r$r;s zChainDataset.__len__) r*r+r,r-rrr6rur;rsr#r#rfr$ris c@sleZdZUdZeeed<eeed<eeeeddddZ dd Z e ee ed d d Z d dZ dS)rz Subset of a dataset at specified indices. Args: dataset (Dataset): The whole Dataset indices (sequence): Indices in the whole set selected for subset rDrPN)rDrPrcCs||_||_dSr'rDrP)r!rDrPr#r#r$r6szSubset.__init__cs2t|tr"jfdd|DSjj|S)Ncsg|]}j|qSr#rO)r1ir:r#r$rSsz&Subset.__getitem__..)rLrGrDrP)r!r[r#r:r$r%s zSubset.__getitem__)rPrcsBttjddr,jfdd|DSfdd|DSdS)NrTcsg|]}j|qSr#rOr1r[r:r#r$rSsz'Subset.__getitems__..csg|]}jj|qSr#rwryr:r#r$rSs)rVrWrDrT)r!rPr#r:r$rTszSubset.__getitems__cCs t|jSr')rBrPr:r#r#r$r;szSubset.__len__)r*r+r,r-rrr<rror6r%rrTr;r#r#r#r$rs   )rDlengths generatorrc s6tt|drt|dkrg}t|D]H\}}|dks@|dkrPtd|dttt|}||q(tt|}t |D] }|t|}||d7<q|}t|D]"\}} | dkrt d|dqt|tkrtdt t||d ttt|}fd d tt||DS) a Randomly split a dataset into non-overlapping new datasets of given lengths. If a list of fractions that sum up to 1 is given, the lengths will be computed automatically as floor(frac * len(dataset)) for each fraction provided. After computing the lengths, if there are any remainders, 1 count will be distributed in round-robin fashion to the lengths until there are no remainders left. Optionally fix the generator for reproducible results, e.g.: Example: >>> # xdoctest: +SKIP >>> generator1 = torch.Generator().manual_seed(42) >>> generator2 = torch.Generator().manual_seed(42) >>> random_split(range(10), [3, 7], generator=generator1) >>> random_split(range(30), [0.3, 0.3, 0.4], generator=generator2) Args: dataset (Dataset): Dataset to be split lengths (sequence): lengths or fractions of splits to be produced generator (Generator): Generator used for the random permutation. rjrzFraction at index z is not between 0 and 1zLength of split at index z- is 0. This might result in an empty dataset.zDSum of input lengths does not equal the length of the input dataset!)r{cs&g|]\}}t|||qSr#)r)r1offsetlengthrwr#r$rSsz random_split..)mathisclosesum enumeraterErofloorrBrYrangewarningswarnrtolistrrrX itertools accumulate) rDrzr{Zsubset_lengthsrxfracZn_items_in_split remainderZ idx_to_add_atr}r#rwr$rs8     )(rkrr~rtypingrrrrrrrr r r typing_extensionsr Ztorchr rrr__all__rrstrZT_dictZT_tuplerrrrrrrrrofloatrr#r#r#r$sF0     "tW9 $