fTh(SrSSKrSSKJrJr SSKJr SSKJr SSK J r J r J r SSK JrJrJrJr SS KJr S S KJr \R,"\5r"S S \ SS9r"SS\ 5rS/rg)zq Processor class for InstructBLIP. Largely copy of Blip2Processor with addition of a tokenizer for the Q-Former. N)ListUnion) BatchFeature) ImageInput)ProcessingKwargsProcessorMixinUnpack) AddedToken BatchEncodingPreTokenizedInput TextInput)logging) AutoTokenizerc 2\rSrSrSSSSSSSSSS. 0S.rSrg) InstructBlipProcessorKwargs&TFr) add_special_tokenspaddingstridereturn_overflowing_tokensreturn_special_tokens_maskreturn_offsets_mappingreturn_token_type_ids return_lengthverbose) text_kwargs images_kwargsN)__name__ __module__ __qualname____firstlineno__ _defaults__static_attributes__r p/var/www/auris/envauris/lib/python3.13/site-packages/transformers/models/instructblip/processing_instructblip.pyrr&s0#').*/&+%*"   Ir'rF)totalc ^\rSrSrSr/SQrS/rSrSrSr SU4Sjjr SS\ S \ \ \\\ \\4S \\S \4S jjrS rSr\S5rU4Sjr\U4Sj5rSrU=r$)InstructBlipProcessor7aX Constructs an InstructBLIP processor which wraps a BLIP image processor and a LLaMa/T5 tokenizer into a single processor. [`InstructBlipProcessor`] offers all the functionalities of [`BlipImageProcessor`] and [`AutoTokenizer`]. See the docstring of [`~BlipProcessor.__call__`] and [`~BlipProcessor.decode`] for more information. Args: image_processor (`BlipImageProcessor`): An instance of [`BlipImageProcessor`]. The image processor is a required input. tokenizer (`AutoTokenizer`): An instance of ['PreTrainedTokenizer`]. The tokenizer is a required input. qformer_tokenizer (`AutoTokenizer`): An instance of ['PreTrainedTokenizer`]. The Q-Former tokenizer is a required input. num_query_tokens (`int`, *optional*):" Number of tokens used by the Qformer as queries, should be same as in model's config. )image_processor tokenizerqformer_tokenizernum_query_tokens)BlipImageProcessorBlipImageProcessorFastrc >[US5(d,[SSSS9UlURUR/SS9 OURUlX@l[ TU]XU5 g)N image_tokenzFT) normalizedspecial)special_tokens)hasattrr r4 add_tokensr0super__init__)selfr-r.r/r0kwargs __class__s r(r;InstructBlipProcessor.__init__Psay-00))tTD   $"2"2!3D I(44D  0 5FGr'imagestextr=returnc UcUc [S5eUR"[4SURR0UD6n[ 5nUGb[ U[5(aU/nO8[ U[5(d#[ US[5(d [S5eUSRSS5nUR"U40USDSS0D6n XSS'URbyUbv0n URRUR-n URU /[U5-SSS 9n U H,n [XX5VVs/sH upX-PM snnX'M. OU n Ub[R!S 5 [#XS 9n UR%U 5 UR&"U40USD6nURS 5US 'URS5US'Ub'UR("U40USD6nUR%U5 U$s snnf)aP This method uses [`BlipImageProcessor.__call__`] method to prepare image(s) for the model, and [`BertTokenizerFast.__call__`] to prepare text for the model. Please refer to the docstring of the above two methods for more information. Args: images (`ImageInput`): The image or batch of images to be prepared. Each image can be a PIL image, NumPy array or PyTorch tensor. Both channels-first and channels-last formats are supported. text (`TextInput`, `PreTokenizedInput`, `List[TextInput]`, `List[PreTokenizedInput]`): The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set `is_split_into_words=True` (to lift the ambiguity with a batch of sequences). Nz,You have to specify at least images or text.tokenizer_init_kwargsrzAInvalid input text. Please provide a string, or a list of stringsrreturn_tensorsF)rrEaAExpanding inputs for image tokens in InstructBLIP should be done in processing. Please follow instruction here (https://gist.github.com/zucchini-nlp/e9f20b054fa322f84ac9311d9ab67042) to update your InstructBLIP model. Using processors without these attributes in the config is deprecated and will throw an error in v4.50.) tensor_type input_idsqformer_input_idsattention_maskqformer_attention_maskr) ValueError _merge_kwargsrr. init_kwargsr isinstancestrlistpopr0r4contentlenziplogger warning_oncer updater/r-)r<r@rAaudiovideosr= output_kwargsencodingrE_text_encoding text_encoding image_tokensimage_token_encodingk img_encoding txt_encodingqformer_text_encodingimage_encodings r(__call__InstructBlipProcessor.__call__Ys1, >dlKL L** ' "&.."<"<   >  $$$vd++JtAw4L4L !dee+=9==>NPTUN!^^DfM-4PfaefN=K- ()9 :$$0V5G " #//77$:O:OO '+~~!NSY.5Y](6($(A;>>R>UWeWh:i(:i6L%3:i(M$( !/ %''B*-TM OOM *$($:$:4$`=Q^C_$` !,A,E,Ek,RH( )1F1J1JK[1\H- .  !11&[M/!% @ @&*&:&:&L&L#DMM"7"UVWWr'c >[RRU5(a[SUS35e[R"USS9 [RR US5nUR RU5 SUR;nU(aURRS5 [TU]"U40UD6nU(aU=RS/- slU$)NzProvided path (z#) should be a directory, not a fileT)exist_okr/) ospathisfilerKmakedirsjoinr/save_pretrained attributesremover:)r<save_directoryr=qformer_tokenizer_pathqformer_presentoutputsr>s r(r~%InstructBlipProcessor.save_pretraineds 77>>. ) )~.>>abc c NT2!#n>Q!R ../EF.@  OO " "#6 7').CFC  OO 34 4Or'c >[TU]"U40UD6n[U[5(aUSn[R"USS9nXClU$)Nrr/) subfolder)r:from_pretrainedrNtuplerr/)clspretrained_model_name_or_pathr= processorr/r>s r(r%InstructBlipProcessor.from_pretrainedsQG+,ITVT  i ' '!! I)99:Wcvw&7#r')r4r0rp)NNNN)r!r"r#r$__doc__r valid_kwargsimage_processor_classtokenizer_classqformer_tokenizer_classr;rrrr rr rrrerhrmpropertyrqr~ classmethodrr& __classcell__)r>s@r(r+r+7s$GJ&'LL%O-H"^b MMI0$y/4HYCZZ[M 45 M M`<6XX &r'r+)rrytypingrrimage_processing_utilsr image_utilsrprocessing_utilsrr r tokenization_utils_baser r r rutilsrautor get_loggerr!rUrr+__all__r r'r(rsj 2%HH     H %"2%"eNeP # #r'