o ¾e¦iMã@sfdZddlmZmZddlmZmZmZddlm Z ddl m Z m Z m Z e Gdd„de ƒƒZdgZd S) z Processor class for Nougat. é)ÚOptionalÚUnion)ÚPreTokenizedInputÚ TextInputÚTruncationStrategyé)ÚProcessorMixin)ÚPaddingStrategyÚ TensorTypeÚauto_docstringcBsÄeZdZ‡fdd„Ze                                  d.dedBd edBd eeefdBd d d edBdedBdedBdedBdee BdBdedBde e e BdBde e e BdBde dde edfdBde eBe e Be eBdBde eBe e Be eBdBde eBe e Be eBdBdedeeBeBdeeBeBdBdedBded ed!edBd"eeBdBd#edBd$edBd%ed&ed'ed(ed)ef@d*d+„ƒZd,d-„Z‡ZS)/ÚNougatProcessorcstƒ ||¡dS)N)ÚsuperÚ__init__)ÚselfÚimage_processorÚ tokenizer©Ú __class__©új/home/ubuntu/transcripts/venv/lib/python3.10/site-packages/transformers/models/nougat/processing_nougat.pyrszNougatProcessor.__init__NÚchannels_firstTFrÚdo_crop_marginÚ do_resizeÚsizeÚresampleÚPILImageResamplingÚ do_thumbnailÚdo_align_long_axisÚdo_padÚ do_rescaleÚrescale_factorÚ do_normalizeÚ image_meanÚ image_stdÚ data_formatÚChannelDimensionÚinput_data_formatÚ text_pairÚ text_targetÚtext_pair_targetÚadd_special_tokensÚpaddingÚ truncationÚ max_lengthÚstrideÚis_split_into_wordsÚpad_to_multiple_ofÚreturn_tensorsÚreturn_token_type_idsÚreturn_attention_maskÚreturn_overflowing_tokensÚreturn_special_tokens_maskÚreturn_offsets_mappingÚ return_lengthÚverbosec#%Csø|dur |dur tdƒ‚|dur'|j|f||||||| | | | | ||||dœŽ}#|durh|j|fid|“d|“d|“d|“d|“d |“d |“d |“d |“d |“d|“d|“d|“d|“d|“d| “d|!“d|"“Ž}$|durn|#S|durt|$S|$d|#d<|#S)ag do_crop_margin (`bool`, *optional*): Whether to automatically crop white margins from document images. When enabled, the processor detects and removes white space around the edges of document pages, which is useful for processing scanned documents or PDFs with large margins. do_thumbnail (`bool`, *optional*): Whether to create a thumbnail version of the image. When enabled, a smaller version of the image is generated alongside the main processed image, which can be useful for preview or faster processing. do_align_long_axis (`bool`, *optional*): Whether to automatically align images so that the longer axis is horizontal. When enabled, portrait images are rotated to landscape orientation, which is typically better for document processing tasks. NzBYou need to specify either an `images` or `text` input to process.)rrrrrrrrr r!r"r#r1r$r&r'r(r)r*r+r,r-r.r/r0r1r2r3r4r5r6r7r8Ú input_idsÚlabels)Ú ValueErrorrr)%rÚimagesÚtextrrrrrrrrr r!r"r#r$r&r'r(r)r*r+r,r-r.r/r0r1r2r3r4r5r6r7r8ÚinputsÚ encodingsrrrÚ__call__sˆ2ÿðÿþýüûúùø ÷ ö õ ô óòñðïîí zNougatProcessor.__call__cOs|jj|i|¤ŽS)zÊ This method forwards all its arguments to NougatTokenizer's [`~PreTrainedTokenizer.post_process_generation`]. Please refer to the docstring of this method for more information. )rÚpost_process_generation)rÚargsÚkwargsrrrrA†sz'NougatProcessor.post_process_generation)"NNNNNNNNNNNNNNrNNNNTFNNrFNNNNFFFFT)Ú__name__Ú __module__Ú __qualname__rr ÚboolÚdictÚstrÚintÚfloatÚlistrrrrr rr r@rAÚ __classcell__rrrrr sÎ Ýüûúùø ÷ ö õ ô óòñðïîíìë êéèçæå äãâá à!ß"Þ# Ýfr N)Ú__doc__ÚtypingrrÚ$transformers.tokenization_utils_baserrrÚprocessing_utilsrÚutilsr r r r Ú__all__rrrrÚs   s