fTh?i.SSKrSSKrSSKJrJrJrJr SSKrSSK J r SSK J r J r JrJr SSKJrJrJrJr \"5(a SSKrSSKJr \"5(a SSKrSS KJr "S S \5r"S S \ 5r\ "\"SS9S5"SS\55r\rg)N)ListOptionalTupleUnion)BasicTokenizer) ExplicitEnumadd_end_docstringsis_tf_availableis_torch_available)ArgumentHandler ChunkPipelineDatasetbuild_pipeline_init_args)/TF_MODEL_FOR_TOKEN_CLASSIFICATION_MAPPING_NAMES),MODEL_FOR_TOKEN_CLASSIFICATION_MAPPING_NAMESc6\rSrSrSrS\\\\44SjrSr g)"TokenClassificationArgumentHandlerz- Handles arguments for token classification. inputsc UbA[U[[45(a&[U5S:a[U5n[U5nOe[U[5(aU/nSnOJ[ b[U[ 5(d[U[ R5(aUS4$[S5eURS5nU(aJ[U[5(a[US[5(aU/n[U5U:wa [S5eX4$)Nrr zAt least one input is required.offset_mappingz;offset_mapping should have the same batch size as the input) isinstancelisttuplelenstrrtypes GeneratorType ValueErrorget)selfrkwargs batch_sizers c/var/www/auris/envauris/lib/python3.13/site-packages/transformers/pipelines/token_classification.py__call__+TokenClassificationArgumentHandler.__call__ s  *VdE]"C"CF VW&\FVJ  $ $XFJ  Z%@%@JvW\WjWjDkDk4< >? ?$45 .$//J~a?PRW4X4X"0!1>"j0 !^__%%N) __name__ __module__ __qualname____firstlineno____doc__rrrr'__static_attributes__r*r)r&rrs &uS$s)^4&r)rc,\rSrSrSrSrSrSrSrSr Sr g ) AggregationStrategy5zDAll the valid aggregation strategies for TokenClassificationPipelinenonesimplefirstaveragemaxr*N) r+r,r-r.r/NONESIMPLEFIRSTAVERAGEMAXr0r*r)r&r2r25sN D F EG Cr)r2T) has_tokenizera ignore_labels (`List[str]`, defaults to `["O"]`): A list of labels to ignore. grouped_entities (`bool`, *optional*, defaults to `False`): DEPRECATED, use `aggregation_strategy` instead. Whether or not to group the tokens corresponding to the same entity together in the predictions or not. stride (`int`, *optional*): If stride is provided, the pipeline is applied on all the text. The text is split into chunks of size model_max_length. Works only with fast tokenizers and `aggregation_strategy` different from `NONE`. The value of this argument defines the number of overlapping tokens between chunks. In other words, the model will shift forward by `tokenizer.model_max_length - stride` tokens each step. aggregation_strategy (`str`, *optional*, defaults to `"none"`): The strategy to fuse (or not) tokens based on the model prediction. - "none" : Will simply not do any aggregation and simply return raw results from the model - "simple" : Will attempt to group entities following the default schema. (A, B-TAG), (B, I-TAG), (C, I-TAG), (D, B-TAG2) (E, B-TAG2) will end up being [{"word": ABC, "entity": "TAG"}, {"word": "D", "entity": "TAG2"}, {"word": "E", "entity": "TAG2"}] Notice that two consecutive B tags will end up as different entities. On word based languages, we might end up splitting words undesirably : Imagine Microsoft being tagged as [{"word": "Micro", "entity": "ENTERPRISE"}, {"word": "soft", "entity": "NAME"}]. Look for FIRST, MAX, AVERAGE for ways to mitigate that and disambiguate words (on languages that support that meaning, which is basically tokens separated by a space). These mitigations will only work on real words, "New york" might still be tagged with two different entities. - "first" : (works only on word based models) Will use the `SIMPLE` strategy except that words, cannot end up with different tags. Words will simply use the tag of the first token of the word when there is ambiguity. - "average" : (works only on word based models) Will use the `SIMPLE` strategy except that words, cannot end up with different tags. scores will be averaged first across tokens, and then the maximum label is applied. - "max" : (works only on word based models) Will use the `SIMPLE` strategy except that words, cannot end up with different tags. Word entity will simply be the token with the maximum score.c(^\rSrSrSrSr\"54U4SjjrS"S\\ S\\ S\\ S \\ \ \ \ 4S \\ 4 S jjrS \\\ \44U4S jjrS#SjrSr\ R(S4SjrSrS\S\R0S\R0S \\ \ \ \ 4S\R0S\ S\ \4SjrS\ \S\ S\ \4SjrS\ \S\ S\4SjrS\ \S\ S\ \4SjrS\ \S\4SjrS\S\ \\44SjrS\ \S\ \4S jr S!r!U=r"$)$TokenClassificationPipeline?u Named Entity Recognition pipeline using any `ModelForTokenClassification`. See the [named entity recognition examples](../task_summary#named-entity-recognition) for more information. Example: ```python >>> from transformers import pipeline >>> token_classifier = pipeline(model="Jean-Baptiste/camembert-ner", aggregation_strategy="simple") >>> sentence = "Je m'appelle jean-baptiste et je vis à montréal" >>> tokens = token_classifier(sentence) >>> tokens [{'entity_group': 'PER', 'score': 0.9931, 'word': 'jean-baptiste', 'start': 12, 'end': 26}, {'entity_group': 'LOC', 'score': 0.998, 'word': 'montréal', 'start': 38, 'end': 47}] >>> token = tokens[0] >>> # Start and end provide an easy way to highlight words in the original text. >>> sentence[token["start"] : token["end"]] ' jean-baptiste' >>> # Some models use the same idea to do part of speech. >>> syntaxer = pipeline(model="vblagoje/bert-english-uncased-finetuned-pos", aggregation_strategy="simple") >>> syntaxer("My name is Sarah and I live in London") [{'entity_group': 'PRON', 'score': 0.999, 'word': 'my', 'start': 0, 'end': 2}, {'entity_group': 'NOUN', 'score': 0.997, 'word': 'name', 'start': 3, 'end': 7}, {'entity_group': 'AUX', 'score': 0.994, 'word': 'is', 'start': 8, 'end': 10}, {'entity_group': 'PROPN', 'score': 0.999, 'word': 'sarah', 'start': 11, 'end': 16}, {'entity_group': 'CCONJ', 'score': 0.999, 'word': 'and', 'start': 17, 'end': 20}, {'entity_group': 'PRON', 'score': 0.999, 'word': 'i', 'start': 21, 'end': 22}, {'entity_group': 'VERB', 'score': 0.998, 'word': 'live', 'start': 23, 'end': 27}, {'entity_group': 'ADP', 'score': 0.999, 'word': 'in', 'start': 28, 'end': 30}, {'entity_group': 'PROPN', 'score': 0.999, 'word': 'london', 'start': 31, 'end': 37}] ``` Learn more about the basics of using a pipeline in the [pipeline tutorial](../pipeline_tutorial) This token recognition pipeline can currently be loaded from [`pipeline`] using the following task identifier: `"ner"` (for predicting the classes of tokens in a sequence: person, organisation, location or miscellaneous). The models that this pipeline can use are models that have been fine-tuned on a token classification task. See the up-to-date list of available models on [huggingface.co/models](https://huggingface.co/models?filter=token-classification). sequencesc>[TU]"U0UD6 URURS:Xa[O[ 5 [ SS9UlXlg)NtfF) do_lower_case) super__init__check_model_type frameworkrrr_basic_tokenizer _args_parser)r# args_parserargsr$ __class__s r&rG$TokenClassificationPipeline.__init__sL $)&) ~~% <= !/U C'r)Ngrouped_entitiesignore_subwordsaggregation_strategyrstridecZ0nUbXWS'0nUcUbU(aU(a[RnO/U(aU(d[RnO[RnUb[R "SUS35 Ub[R "SUS35 Ub[ U[5(a[UR5nU[R[R[R1;a&URR(d [S5eXHS'UbXS'UbxX`RR:a [S5eU[R:Xa[S US 35eURR(a S S US .n XS 'O [S5eU0U4$)Nrzl`grouped_entities` is deprecated and will be removed in version v5.0.0, defaulted to `aggregation_strategy="z "` instead.zk`ignore_subwords` is deprecated and will be removed in version v5.0.0, defaulted to `aggregation_strategy="z{Slow tokenizers cannot handle subwords. Please set the `aggregation_strategy` option to `"simple"` or use a fast tokenizer.rR ignore_labelszl`stride` must be less than `tokenizer.model_max_length` (or even lower if the tokenizer adds special tokens)zI`stride` was provided to process all the text but `aggregation_strategy="z&"`, please select another one instead.T)return_overflowing_tokenspaddingrStokenizer_paramszm`stride` was provided to process all the text but you're using a slow tokenizer. Please use a fast tokenizer.)r2r;r:r9warningswarnrrupperr=r< tokenizeris_fastr!model_max_length) r#rUrPrQrRrrSpreprocess_paramspostprocess_paramsrXs r&_sanitize_parameters0TokenClassificationPipeline._sanitize_parameterss  %2@. /  '?+FO':'@'@$!/':'A'A$':'?'?$+ //C.DKQ* //C.DKQ +.44':;O;U;U;W'X$$'--/B/F/FH[HcHcde.. >:N5 6  $2? /  888 C$':'?'?? ,--SU >>))59#'"(($ =M&89$8!"&888r)rc f>UR"U40UD6up4U(aXBS'[TU]"U40UD6$)a- Classify each token of the text(s) given as inputs. Args: inputs (`str` or `List[str]`): One or several texts (or one list of texts) for token classification. Return: A list or a list of list of `dict`: Each result comes as a list of dictionaries (one for each token in the corresponding input, or each entity if this pipeline was instantiated with an aggregation_strategy) with the following keys: - **word** (`str`) -- The token/word classified. This is obtained by decoding the selected tokens. If you want to have the exact string in the original sentence, use `start` and `end`. - **score** (`float`) -- The corresponding probability for `entity`. - **entity** (`str`) -- The entity predicted for that token/word (it is named *entity_group* when *aggregation_strategy* is not `"none"`. - **index** (`int`, only present when `aggregation_strategy="none"`) -- The index of the corresponding token in the sentence. - **start** (`int`, *optional*) -- The index of the start of the corresponding entity in the sentence. Only exists if the offsets are available within the tokenizer - **end** (`int`, *optional*) -- The index of the end of the corresponding entity in the sentence. Only exists if the offsets are available within the tokenizer r)rKrFr')r#rr$_inputsrrNs r&r'$TokenClassificationPipeline.__call__s>4#'"3"3F"Ef"E '5# $w1&11r)c +# URS05nURR(aURRS:aSOSnUR"U4URUSURRS.UD6nURSS5 [ US5n[ U5HnURS:Xa=UR5V V s0sHupU [R"XS5_M! n n n O6UR5V V s0sHupXURS5_M n n n UbX+S 'US:XaUOSU S 'XS - :HU S 'U v M gs sn n fs sn n f7f) NrXrTF)return_tensors truncationreturn_special_tokens_maskreturn_offsets_mappingoverflow_to_sample_mapping input_idsrDrsentencer is_last) popr\r^rIr]rrangeitemsrD expand_dims unsqueeze) r#rmrr_rXrhr num_chunksikv model_inputss r& preprocess&TokenClassificationPipeline.preprocesssN,001CRH!^^<<A`A`cdAdTjo   >>!'+#'>>#9#9      /6 ,- z"A~~%GM||~V~tq2>>!$#: :~ V AGPQ4>>!#4 4 P)1?-.346xtL $&'>&9L # #VPsCE(&E;E( E"26E(cNURS5nURSS5nURS5nURS5nURS:XaUR"S 0UD6SnO1UR"S 0UD6n[U[5(aUSOUSnUUUUUS.UE$) Nspecial_tokens_maskrrmrnrDrlogits)r}r|rrmrnr*)rorImodelrdict)r#rxr|rrmrnr}outputs r&_forward$TokenClassificationPipeline._forwards*../DE%))*:DA##J/""9- >>T !ZZ/,/2FZZ/,/F)3FD)A)AVH%vayF#6,      r)c UcS/n/nUGHnURS:XahUSSR[R[R4;a4USSR [R 5R5nOUSSR5nUSSnUSSnUSbUSSOSn USSR5n [R"US S S 9n [R"Xk- 5n XRS S S 9- n URS :Xa%UR5nU bU R5OSn URXxXX5nURX5nUVs/sH5nURS S5U;dMURSS5U;dM3UPM7 nnURU5 GM [!U5nUS:aUR#U5nU$s snf)NOptr}rrmrlrr|T)axiskeepdimsrDentity entity_groupr )rIdtypetorchbfloat16float16tofloat32numpynpr8expsumgather_pre_entities aggregater"extendraggregate_overlapping_entities)r# all_outputsrRrU all_entities model_outputsr}rmrlrr|maxes shifted_expscores pre_entitiesrPrentitiesrts r& postprocess'TokenClassificationPipeline.postprocess+s   EM (M~~%-*A!*D*J*Ju~~_d_l_lNm*m&x0366u}}EKKM&x0399;"1~j1H%k215I6CDT6U6a ./2gk #00E"Fq"I"O"O"Q FF6T:E&&0K ??T?#JJF~~%%OO- ;I;U!5!5!7[_33V=PL $~~lQ /.F::h-]BJJ~t4MI.     )A)B% >>>|LLs0G(G(&G(cH[U5S:XaU$[USS9n/nUSnUHgnUSUSs=::a US:a;O O8USUS- nUSUS- nXV:aUnM;XV:XaUSUS:aUnMPMRMTURU5 UnMi URU5 U$)Nrc US$)Nstartr*)xs r&LTokenClassificationPipeline.aggregate_overlapping_entities..Xs!G*r)keyrendscore)rsortedappend)r#raggregated_entitiesprevious_entityrcurrent_lengthprevious_lengths r&r:TokenClassificationPipeline.aggregate_overlapping_entitiesUs x=A O((<= "1+Fw'6'?S_U=SS!'!@"1%"8?7;S"S!3&,O#66'?_]dMe;e&,O>77IN8KLD)%3%8" !)S11~~-$-NN$4 "),,.#g64>><>>7NN--335PRVDD "%Tc(m!;J,+11+33+//0 ! ]'"+Q!e3hST}W`cdWd>e3eJy~&$..*E*EE#D!&J " &"( J    +a"3br)rcU[R[R1;an/nUHenUSR5nUSUnURR R UUUSUSUSUSS.nURU5 Mg OURX5nU[R:XaU$URU5$)Nrrrrr)rrrrrr) r2r9r:argmaxr~configid2labelraggregate_wordsgroup_entities)r#rrRrr entity_idxrrs r&r%TokenClassificationPipeline.aggregates $7$<$<>Q>X>X#Y YH* '188: "8,Z8"jj//88D"'0&v.'0%e, '+++LOH #6#;#; ;O""8,,r)rc(URRUVs/sHo3SPM sn5nU[R:Xa@USSnUR 5nXVnUR R RUnOU[R:XaH[USS9nUSnUR 5nXVnUR R RUnOU[R:Xav[R"UVs/sHo3SPM sn5n[R"USS9n U R 5n UR R RU nXnO [S5eUUUUSSUS S S .n U $s snfs snf) Nrrrc(USR5$)Nr)r8)rs r&r.s&:J:N:N:Pr)r)rzInvalid aggregation_strategyrrr)rrrrr)r\convert_tokens_to_stringr2r;rr~rrr=r8r<rstacknanmeanr!) r#rrRrrrrr max_entityaverage_scoresr new_entitys r&aggregate_word*TokenClassificationPipeline.aggregate_words|~~66U]7^U]6vU]7^_ #6#<#< <a[*F--/CKEZZ&&//4F !%8%<%< <X+PQJ)F--/CKEZZ&&//4F !%8%@%@ @XXhGhFh/hGHFZZQ7N'..0JZZ&&// ;F".E;< <a[)B<&  78_Hs F FcTU[R[R1;a [S5e/nSnUHKnUcU/nM US(aUR U5 M(UR UR XB55 U/nMM Ub UR UR XB55 U$)z Override tokens from a given word that disagree to force agreement on word boundaries. Example: micro|soft| com|pany| B-ENT I-NAME I-ENT I-ENT will be rewritten with first strategy as microsoft| company| B-ENT I-ENT z;NONE and SIMPLE strategies are invalid for word aggregationNr)r2r9r:r!rr)r#rrR word_entities word_grouprs r&r+TokenClassificationPipeline.aggregate_wordss  $ $  & &$  Z[ [  F!$X  %!!&)$$T%8%8%Z[$X   !  !4!4Z!V Wr)cLUSSRSS5Sn[R"UVs/sHo"SPM sn5nUVs/sHo"SPM nnW[R"U5URR U5USSUSS S .nU$s snfs snf) z Group together the adjacent tokens with the same entity predicted. Args: entities (`dict`): The entities predicted by the pipeline. rr-r rrrrr)rrrrr)splitrrmeanr\r)r#rrrtokensrs r&group_sub_entities.TokenClassificationPipeline.group_sub_entitiess!X&,,S!4R88D8G_8DE/78xV.x8#WWV_NN;;FCa[)B<&  E8s B B! entity_namecURS5(a SnUSSnX#4$URS5(a SnUSSnX#4$SnUnX#4$)NzB-BrzI-I) startswith)r#rbitags r&get_tag#TokenClassificationPipeline.get_tag sk  ! !$ ' 'Bab/Cw # #D ) )Bab/C wBCwr)c/n/nUHnU(dURU5 MURUS5upVURUSS5upxXh:XaUS:waURU5 MjURURU55 U/nM U(a URURU55 U$)z Find and group together the adjacent tokens with the same entity predicted. Args: entities (`dict`): The entities predicted by the pipeline. rrr)rrr) r#r entity_groupsentity_group_disaggrrrlast_bilast_tags r&r*TokenClassificationPipeline.group_entitiess  F&#**62 ll6(#34GB $ -@-DX-N O G29#**62$$T%<%<=P%QR'-h#'(   !8!89L!M Nr))rKrJ)NNNNNN)N)#r+r,r-r.r/default_input_namesrrGrboolr2rrrrarrr'ryrr9rrrndarrayrrrrrrrrr0 __classcell__)rNs@r&r@r@?sF"H&#E#G (+/*.>B:> $G9#4.G9"$ G9 '':; G9 !eCHo!67 G9 G9R2uS$s)^42@4 *=Prs //: TS^X&&4,40n!D\-\E!D\~* r)