fTh2'SrSSKJrJrJr SSKJr SSKJr SSK J r SSK J r J r JrJr SSKJrJr \R&"\5r"S S \S S 9r"S S\ S S 9r"SS\ 5rS/rg)z Processor class for UDOP. )ListOptionalUnion)logging) BatchFeature) ImageInput)ProcessingKwargsProcessorMixin TextKwargsUnpack)PreTokenizedInput TextInputct\rSrSr%\\\\\\\4\S'\\\\\\\\4\S'Sr g)UdopTextKwargs word_labelsboxesN) __name__ __module__ __qualname____firstlineno__rrrint__annotations____static_attributes__r`/var/www/auris/envauris/lib/python3.13/site-packages/transformers/models/udop/processing_udop.pyrr sH%S 4S ? :;<< d3i$tDI"77 88rrF)totalc >\rSrSr%\\S'SSSSSSSSSS. 0S.rSrg ) UdopProcessorKwargs% text_kwargsTFr) add_special_tokenspadding truncationstridereturn_overflowing_tokensreturn_special_tokens_maskreturn_offsets_mapping return_lengthverbose)r# images_kwargsrN)rrrrrr _defaultsrrrrr!r!%s6#').*/&+"   Irr!c ^\rSrSrSrSS/rSrSrS/rU4Sjr SS S S .S \ \ S \ \ \\\ \\4S \\S\4SjjjrSrSrSr\S5rSrU=r$) UdopProcessor7a Constructs a UDOP processor which combines a LayoutLMv3 image processor and a UDOP tokenizer into a single processor. [`UdopProcessor`] offers all the functionalities you need to prepare data for the model. It first uses [`LayoutLMv3ImageProcessor`] to resize, rescale and normalize document images, and optionally applies OCR to get words and normalized bounding boxes. These are then provided to [`UdopTokenizer`] or [`UdopTokenizerFast`], which turns the words and bounding boxes into token-level `input_ids`, `attention_mask`, `token_type_ids`, `bbox`. Optionally, one can provide integer `word_labels`, which are turned into token-level `labels` for token classification tasks (such as FUNSD, CORD). Additionally, it also supports passing `text_target` and `text_pair_target` to the tokenizer, which can be used to prepare labels for language modeling tasks. Args: image_processor (`LayoutLMv3ImageProcessor`): An instance of [`LayoutLMv3ImageProcessor`]. The image processor is a required input. tokenizer (`UdopTokenizer` or `UdopTokenizerFast`): An instance of [`UdopTokenizer`] or [`UdopTokenizerFast`]. The tokenizer is a required input. image_processor tokenizerLayoutLMv3ImageProcessor) UdopTokenizerUdopTokenizerFast text_pairc$>[TU]X5 g)N)super__init__)selfr2r3 __class__s rr:UdopProcessor.__init__Ss 4rN)audiovideosimagestextkwargsreturncUR"[4SURR0UDUR"U6D6nUSR SS5nUSR SS5n USR SS5n USR SS5n USR S S5n USR S S5n URR(aUb [S 5eURR(aU b [S 5eU (aU (d [S 5eU bUR"S0USD6$UR"SSU0USD6nUR SS5nUR SS5nUSR S S5 USR SS5 XSS'UbUOUUSS'XSS'Ub=URR(a"U c[U[5(aU/nXSS'UR"SSUbUOU0USD6nU SLaURUSUS5US'URU5 U$)a. This method first forwards the `images` argument to [`~UdopImageProcessor.__call__`]. In case [`UdopImageProcessor`] was initialized with `apply_ocr` set to `True`, it passes the obtained words and bounding boxes along with the additional arguments to [`~UdopTokenizer.__call__`] and returns the output, together with the prepared `pixel_values`. In case [`UdopImageProcessor`] was initialized with `apply_ocr` set to `False`, it passes the words (`text`/``text_pair`) and `boxes` specified by the user along with the additional arguments to [`~UdopTokenizer.__call__`] and returns the output, together with the prepared `pixel_values`. Alternatively, one can pass `text_target` and `text_pair_target` to prepare the targets of UDOP. Please refer to the docstring of the above two methods for more information. tokenizer_init_kwargsr#rNrr7r(Fr* text_targetzdYou cannot provide bounding boxes if you initialized the image processor with apply_ocr set to True.zaYou cannot provide word labels if you initialized the image processor with apply_ocr set to True.zKYou cannot return overflowing tokens without returning the offsets mapping.r@r-wordstext_pair_targetrAT pixel_valuesoverflow_to_sample_mappingr) _merge_kwargsr!r3 init_kwargs'prepare_and_validate_optional_call_argspopgetr2 apply_ocr ValueError isinstancestrget_overflowing_imagesupdate)r;r@rAr>r?argsrB output_kwargsrrr7r(r*rFfeaturesfeatures_wordsfeatures_boxesencoded_inputss r__call__UdopProcessor.__call__Vs8**  "&.."<"<  ::DA  m,00$?#M266}dK !-044[$G $1-$@$D$DE`bg$h!!.}!=!A!ABZ\a!b#M266}dK    ) )u/@v     ) ){/Fs  %-Cjk k  ">> . ++\6\]?=[\H%\\'48N%\\'48N - ( , ,]D A - ( , ,-? F8A- ( 5=B=NETbM- ( 1:E- ( 7D$8$8$B$ByGXdC(( 6D .N )D0+/+F+F^,n=Y.Z,( OON +Orc/nUHnURX5 M [U5[U5:wa#[S[U5S[U535eU$)Nz`Expected length of images to be the same as the length of `overflow_to_sample_mapping`, but got z and )appendlenrQ)r;r@rJimages_with_overflow sample_idxs rrT$UdopProcessor.get_overflowing_imagessr!4J ' '(: ;5 # $,F(G G,-.eC8R4S3TV  $#rc:URR"U0UD6$)z This method forwards all its arguments to PreTrainedTokenizer's [`~PreTrainedTokenizer.batch_decode`]. Please refer to the docstring of this method for more information. )r3 batch_decoder;rVrBs rreUdopProcessor.batch_decodes ~~**D;F;;rc:URR"U0UD6$)z This method forwards all its arguments to PreTrainedTokenizer's [`~PreTrainedTokenizer.decode`]. Please refer to the docstring of this method for more information. )r3decoderfs rriUdopProcessor.decodes ~~$$d5f55rc /SQ$)N)rI input_idsbboxattention_maskr)r;s rmodel_input_namesUdopProcessor.model_input_namessFFrr)NN)rrrr__doc__ attributesimage_processor_classtokenizer_classoptional_call_argsr:rr rrrrr r!rr\rTreripropertyror __classcell__)r<s@rr0r07s*$[1J6rsn)( 2%TTC   H %9Zu9 *%$^GN^GB  r