sgo> ,UdZddlZddlZddlZddlZddlmZmZmZm Z m Z m Z ddl Z ejdZdZdZdddeee e e e d ffeffd Zdddeee e e e d ffeffd Zdddeee e e e d ffeffd Zdddeee e e e d ffeffd Zdddeee e e e d ffeffdZdddeee e e e d ffeffdZdddeee e e e d ffeffdZe j0eiZee e e e d ffefed<ej6e5ddlZeeej><eeej@ejBejDf<dddeee#<eee$<eee%<eee&<dZ'y#1swY!xYw)aLContains definitions of the methods used by the _BaseDataLoaderIter workers. These methods are used to collate samples fetched from dataset into Tensor(s). These **needs** to be in global scope since Py2 doesn't support serializing static methods. `default_collate` and `default_convert` are exposed to users via 'dataloader.py'. N)CallableDictOptionalTupleTypeUnionz[SaUO]c t|}t|tjr|S|jdk(rm|j dk7r^|j dk7rO|j dk(r+t j|jj|Stj|St|tjjr t|tjjrCtj|}|j!|Dcic]}|t#||c}|S||Dcic]}|t#||c}St|t&rt)|dr |d|DSt|t&r|Dcgc] }t#|c}St|tjj*rt|tt,fs| t|tjj.r8tj|}t1|D]\}}t#|||<|S||Dcgc] }t#|c}S|Scc}wcc}w#t$$r&|Dcic]}|t#||ncc}wc}cYSwxYwcc}wcc}w#t$$r"|Dcgc] }t#|ncc}wc}cYSwxYw)a Convert each NumPy array element into a :class:`torch.Tensor`. If the input is a `Sequence`, `Collection`, or `Mapping`, it tries to convert each element inside to a :class:`torch.Tensor`. If the input is not an NumPy array, it is left unchanged. This is used as the default function for collation when both `batch_sampler` and `batch_size` are NOT defined in :class:`~torch.utils.data.DataLoader`. The general input type to output type mapping is similar to that of :func:`~torch.utils.data.default_collate`. See the description there for more details. Args: data: a single data point to be converted Examples: >>> # xdoctest: +SKIP >>> # Example with `int` >>> default_convert(0) 0 >>> # Example with NumPy array >>> default_convert(np.array([0, 1])) tensor([0, 1]) >>> # Example with NamedTuple >>> Point = namedtuple('Point', ['x', 'y']) >>> default_convert(Point(0, 0)) Point(x=0, y=0) >>> default_convert(Point(np.array(0), np.array(0))) Point(x=tensor(0), y=tensor(0)) >>> # Example with List >>> default_convert([np.array([0, 1]), np.array([2, 3])]) [tensor([0, 1]), tensor([2, 3])] numpystr_string_ndarray_fieldsc32K|]}t|ywN)default_convert).0ds R/var/www/html/venv/lib/python3.12/site-packages/torch/utils/data/_utils/collate.py z"default_convert..Ws >? $ N#c?49#==NO  T!Rc#tCy'A"A!RSS D% WT9%=>> def collate_tensor_fn(batch, *, collate_fn_map): ... # Extend this function to handle batch of tensors ... return torch.stack(batch, 0) >>> def custom_collate(batch): ... collate_map = {torch.Tensor: collate_tensor_fn} ... return collate(batch, collate_fn_map=collate_map) >>> # Extend `default_collate` by in-place modifying `default_collate_fn_map` >>> default_collate_fn_map.update({torch.Tensor: collate_tensor_fn}) Note: Each collate function requires a positional argument for batch and a keyword argument for the dictionary of collate functions as `collate_fn_map`. rr3rc38K|]}t|yw)r3N)collate)rsamplesr4s rrzcollate..s#??sc3:K|]}t|k(ywr)len)relem elem_sizes rrzcollate..s9d3t9 )9sz5each element in list of batch should be of equal size)rrr!r"r#r$r%r&r7r'r(r)zipr*iterr:nextall RuntimeErrorlistr,r-default_collate_err_msg_formatformat) batchr4r;r/ collate_typer0r1rit transposedr8r2r<s ` @rr7r7vs8@ 8DT I!  &,>),U>R R* L$ -3~l3.  $ //0 $  > >? $ $(  W-23QsV3N  $(  W-23QsV3N D% WT9%="E{  D+//22 3 %[RM 9b99VW W#u+& dE " *?   dKOO$C$CD!IIdOE&/ &;S 7#*7>#RaS L$,6 '$GNK 299)D EEG44   We4af44^TT  .    $.GNC s-A J6J  J J  J&J-J < J J JK(*AK2K2K-"K2J JJ JK%+K: KK K%$K%-K22 L?L LLc|d}d}|jr td|jtjtj tj tjtjhvr tdtjjjtd|D}|jj||j}|j!|j"t%|gt'|j)}tj*|d|S)NrzBatches of nested tensors are not currently supported by the default collate_fn; please provide a custom collate_fn to handle them appropriately.zBatches of sparse tensors are not currently supported by the default collate_fn; please provide a custom collate_fn to handle them appropriately.c3<K|]}|jywr)numel)rxs rrz$collate_tensor_fn.. s-!AGGI-s)device)out) is_nestedrAlayoutr sparse_coo sparse_csr sparse_bsr sparse_csc sparse_bscutilsr.get_worker_infosum_typed_storage _new_sharedrMnewresize_r:rBsizestack)rEr4r;rNrKstorages rcollate_tensor_fnr`s 8D C ~~ O   {{       O   {{'')5-u--%%'33E$++3N'dhhw''E GT$))+5FG ;;uaS ))c|d}tj|jj(t t j |jt|Dcgc]}tj|c}|Scc}w)Nrr3) rrrrr'rCrDr7rr )rEr4r;bs rcollate_numpy_array_fnrdse 8D&&tzz~~6B6==djjIJJ 61EOOA&6~ VV6s Bc,tj|Sr)rr rEr4s rcollate_numpy_scalar_fnrg s ??5 !!racLtj|tjS)N)r)rtensorfloat64rfs rcollate_float_fnrk(s <<U]] 33rac,tj|Sr)rrirfs rcollate_int_fnrm0s << rac|Srrfs rcollate_str_fnrp8s Lradefault_collate_fn_mapc$t|tS)a Take in a batch of data and put the elements within the batch into a tensor with an additional outer dimension - batch size. The exact output type can be a :class:`torch.Tensor`, a `Sequence` of :class:`torch.Tensor`, a Collection of :class:`torch.Tensor`, or left unchanged, depending on the input type. This is used as the default function for collation when `batch_size` or `batch_sampler` is defined in :class:`~torch.utils.data.DataLoader`. Here is the general input type (based on the type of the element within the batch) to output type mapping: * :class:`torch.Tensor` -> :class:`torch.Tensor` (with an added outer dimension batch size) * NumPy Arrays -> :class:`torch.Tensor` * `float` -> :class:`torch.Tensor` * `int` -> :class:`torch.Tensor` * `str` -> `str` (unchanged) * `bytes` -> `bytes` (unchanged) * `Mapping[K, V_i]` -> `Mapping[K, default_collate([V_1, V_2, ...])]` * `NamedTuple[V1_i, V2_i, ...]` -> `NamedTuple[default_collate([V1_1, V1_2, ...]), default_collate([V2_1, V2_2, ...]), ...]` * `Sequence[V1_i, V2_i, ...]` -> `Sequence[default_collate([V1_1, V1_2, ...]), default_collate([V2_1, V2_2, ...]), ...]` Args: batch: a single batch to be collated Examples: >>> # xdoctest: +SKIP >>> # Example with a batch of `int`s: >>> default_collate([0, 1, 2, 3]) tensor([0, 1, 2, 3]) >>> # Example with a batch of `str`s: >>> default_collate(['a', 'b', 'c']) ['a', 'b', 'c'] >>> # Example with `Map` inside the batch: >>> default_collate([{'A': 0, 'B': 1}, {'A': 100, 'B': 100}]) {'A': tensor([ 0, 100]), 'B': tensor([ 1, 100])} >>> # Example with `NamedTuple` inside the batch: >>> Point = namedtuple('Point', ['x', 'y']) >>> default_collate([Point(0, 0), Point(1, 1)]) Point(x=tensor([0, 1]), y=tensor([0, 1])) >>> # Example with `Tuple` inside the batch: >>> default_collate([(0, 1), (2, 3)]) [tensor([0, 2]), tensor([1, 3])] >>> # Example with `List` inside the batch: >>> default_collate([[0, 1], [2, 3]]) [tensor([0, 2]), tensor([1, 3])] >>> # Two options to extend `default_collate` to handle specific type >>> # Option 1: Write custom collate function and invoke `default_collate` >>> def custom_collate(batch): ... elem = batch[0] ... if isinstance(elem, CustomType): # Some custom condition ... return ... ... else: # Fall back to `default_collate` ... return default_collate(batch) >>> # Option 2: In-place modify `default_collate_fn_map` >>> def collate_customtype_fn(batch, *, collate_fn_map=None): ... return ... >>> default_collate_fn_map.update(CustomType, collate_customtype_fn) >>> default_collate(batch) # Handle `CustomType` automatically r3)r7rq)rEs rdefault_collatersQsz 5)? @@ra)(__doc__r! contextlibr%retypingrrrrrrrcompilerrrCr7r`rdrgrkrmrprrq__annotations__suppress ImportErrorr npr bool_numberobject_floatintrr+rsrorarrs ?? &2::i0VtOSzFT%eD#I.>(>"?"IJKzF@OS*T%eD#I.>(>"?"IJK*FOS WT%eD#I.>(>"?"IJK W OS"T%eD#I.>(>"?"IJK"OS4T%eD#I.>(>"?"IJK4OST%eD#I.>(>"?"IJKOST%eD#I.>(>"?"IJK LL#IU4tSy)9#9:HDEZ%X*@2::&AXBHHbii<=X!1u,s,s .u=AXXs 0:F  F