`ijdZddlZddlmZddlmZmZddlmZddl m Z ddl m Z m Z mZdd lmZmZGd d e d ZGdde ZdgZdS)z Processor class for TrOCR. N)contextmanager)OptionalUnion) BatchFeature) ImageInput)ProcessingKwargsProcessorMixinUnpack)PreTokenizedInput TextInputceZdZiZdS)TrOCRProcessorKwargsN)__name__ __module__ __qualname__ _defaults~/home/jaya/work/projects/VOICE-AGENT/VIET/agent-env/lib/python3.11/site-packages/transformers/models/trocr/processing_trocr.pyrrsIIIrrF)totalc eZdZdZddgZdZdZdfd Z ddee d e e e e e e e fd eed efd Zed ZedZedZedZxZS)TrOCRProcessora Constructs a TrOCR processor which wraps a vision image processor and a TrOCR tokenizer into a single processor. [`TrOCRProcessor`] offers all the functionalities of [`ViTImageProcessor`/`DeiTImageProcessor`] and [`RobertaTokenizer`/`XLMRobertaTokenizer`]. See the [`~TrOCRProcessor.__call__`] and [`~TrOCRProcessor.decode`] for more information. Args: image_processor ([`ViTImageProcessor`/`DeiTImageProcessor`], *optional*): An instance of [`ViTImageProcessor`/`DeiTImageProcessor`]. The image processor is a required input. tokenizer ([`RobertaTokenizer`/`XLMRobertaTokenizer`], *optional*): An instance of [`RobertaTokenizer`/`XLMRobertaTokenizer`]. The tokenizer is a required input. image_processor tokenizerAutoImageProcessor AutoTokenizerNc d}d|vr/tjdt|d}||n|}t |||j|_d|_dS)Nfeature_extractorzhThe `feature_extractor` argument is deprecated and will be removed in v5, use `image_processor` instead.F) warningswarn FutureWarningpopsuper__init__rcurrent_processor_in_target_context_manager)selfrrkwargsr __class__s rr%zTrOCRProcessor.__init__4s  & ( ( M    !' +> ? ? -<-H//N_ )444!%!5*/'''rimagestextr)returnc |jr|j|fi|S||td|jtfd|jji|}||j|fi|d}||j|fi|d}||S||S|d|d<|S)a When used in normal mode, this method forwards all its arguments to AutoImageProcessor's [`~AutoImageProcessor.__call__`] and returns its output. If used in the context [`~TrOCRProcessor.as_target_processor`] this method forwards all its arguments to TrOCRTokenizer's [`~TrOCRTokenizer.__call__`]. Please refer to the docstring of the above two methods for more information. NzBYou need to specify either an `images` or `text` input to process.tokenizer_init_kwargs images_kwargs text_kwargs input_idslabels)r'r& ValueError _merge_kwargsrr init_kwargsr) r(r+r,audiovideosr) output_kwargsinputs encodingss r__call__zTrOCRProcessor.__call__Ds  * <)4)&;;F;; ; >dlabb b**   "&."<    )T)&SSM/4RSSF  &tLL}]/KLLI <M ^ (5F8 Mrc&|jj}|dgzS)Nr3)rmodel_input_names)r(image_processor_input_namess rr>z TrOCRProcessor.model_input_namesls&*&:&L#*hZ77rc#Ktjdd|_|j|_dV|j|_d|_dS)z Temporarily sets the tokenizer for processing the input. Useful for encoding the labels when fine-tuning TrOCR. z`as_target_processor` is deprecated and will be removed in v5 of Transformers. You can process your labels by using the argument `text` of the regular `__call__` method (either in the same call as your images inputs, or in a separate call.TNF)r r!r'rr&rr(s ras_target_processorz"TrOCRProcessor.as_target_processorqsW   9   +/'!% !%!5*/'''rcDtjdt|jS)Nzg`feature_extractor_class` is deprecated and will be removed in v5. Use `image_processor_class` instead.)r r!r"image_processor_classrAs rfeature_extractor_classz&TrOCRProcessor.feature_extractor_classs' u    ))rcDtjdt|jS)Nz[`feature_extractor` is deprecated and will be removed in v5. Use `image_processor` instead.)r r!r"rrAs rrz TrOCRProcessor.feature_extractors' i    ##r)NN)NNNN)rrr__doc__ attributesrDtokenizer_classr%rrrr r listr rrr<propertyr>rrBrEr __classcell__)r*s@rrr!s<  $[1J0%O000000$(,^b &&$&I0$y/4HYCZZ[& -. & &&&&P88X8 0 0^ 0**X*$$X$$$$$rr)rGr contextlibrtypingrrimage_processing_utilsr image_utilsrprocessing_utilsr r r tokenization_utils_baser r rr__all__rrrrTs%%%%%%""""""""222222%%%%%%HHHHHHHHHHCCCCCCCC+5n$n$n$n$n$^n$n$n$b  r