241 lines
32 KiB
Plaintext
241 lines
32 KiB
Plaintext
|
Ë
|
|||
|
|
>û gq ã # ó´ — d dl Z d dlmZ d dlmZ d dlmZmZmZm Z m
|
|||
|
|
Z
|
|||
|
|
m Z m Z d dl
|
|||
|
|
Z
|
|||
|
|
d dlmZmZmZmZmZ d dlmZ ddlmZ dd lmZ dd
|
|||
|
|
lmZm Z dd l m Z dd l m Z dd
|
|||
|
|
l!m"Z"m#Z# ddl$m%Z%m&Z&m'Z' ddlm(Z(m)Z) ddl*m+Z+ ddl,m-Z- ddl.m/Z/ ddl0m1Z1 dZ2dZ3 e« ji e3« d Z5 e jl dg d¢dge5g d dddd d id d id d idd d!idddd"œd#ddd$œ¬%« dd&œd'e d(e7ded)e e7 d*e8d+e9d,e9d-e8d.ee e#ge e f d/ee e e# ge e e f d0ee+e8ge f d1e9d2e e d3e9d4e8d5e e: f d6„« Z;d7„ Z< e)jz d!« d8„ « Z> G d9„ d:e1« Z?y);é N)Úislice)ÚPath)ÚAnyÚCallableÚDictÚIterableÚListÚOptionalÚUnion)ÚConfigÚCosineDistanceÚModelÚ OptimizerÚset_dropout_rate)ÚFloats2dé )Úutil)ÚErrors)Ú CandidateÚ
|
|||
|
|
KnowledgeBase)ÚLanguage)ÚScorer)ÚDocÚSpan)ÚExampleÚvalidate_examplesÚvalidate_get_examples)ÚSimpleFrozenListÚregistry)ÚVocabé )ÚEntityLinker_v1)Údeserialize_config)Ú
|
|||
|
|
TrainablePipeTzç
|
|||
|
|
[model]
|
|||
|
|
@architectures = "spacy.EntityLinker.v2"
|
|||
|
|
|
|||
|
|
[model.tok2vec]
|
|||
|
|
@architectures = "spacy.HashEmbedCNN.v2"
|
|||
|
|
pretrained_vectors = null
|
|||
|
|
width = 96
|
|||
|
|
depth = 2
|
|||
|
|
embed_size = 2000
|
|||
|
|
window_size = 1
|
|||
|
|
maxout_pieces = 3
|
|||
|
|
subword_features = true
|
|||
|
|
ÚmodelÚ
|
|||
|
|
entity_linker)zdoc.entsz doc.sentsz
|
|||
|
|
token.ent_iobztoken.ent_typeztoken.ent_kb_idé@ z@misczspacy.CandidateGenerator.v1z spacy.CandidateBatchGenerator.v1zspacy.EmptyKB.v2z@scorersz spacy.entity_linker_scorer.v1)r% Úlabels_discardÚn_sentsÚ
|
|||
|
|
incl_priorÚ incl_contextÚentity_vector_lengthÚget_candidatesÚget_candidates_batchÚgenerate_empty_kbÚ overwriteÚscorerÚ
|
|||
|
|
use_gold_entsÚcandidates_batch_sizeÚ thresholdg ð?)Ú nel_micro_fÚ nel_micro_rÚ nel_micro_p)ÚrequiresÚassignsÚdefault_configÚdefault_score_weights)r4 ÚnlpÚnamer( r) r* r+ r, r- r. r/ r0 r1 r2 r3 r4 c
|
|||
|
|
óÄ — |j j dd« s t | j ||||||||| | ¬« S t | j ||||||||| |
|
|||
|
|
| | |
|
|||
|
|
||¬« S )a¤ Construct an EntityLinker component.
|
|||
|
|
|
|||
|
|
model (Model[List[Doc], Floats2d]): A model that learns document vector
|
|||
|
|
representations. Given a batch of Doc objects, it should return a single
|
|||
|
|
array, with one row per item in the batch.
|
|||
|
|
labels_discard (Iterable[str]): NER labels that will automatically get a "NIL" prediction.
|
|||
|
|
n_sents (int): The number of neighbouring sentences to take into account.
|
|||
|
|
incl_prior (bool): Whether or not to include prior probabilities from the KB in the model.
|
|||
|
|
incl_context (bool): Whether or not to include the local context in the model.
|
|||
|
|
entity_vector_length (int): Size of encoding vectors in the KB.
|
|||
|
|
get_candidates (Callable[[KnowledgeBase, Span], Iterable[Candidate]]): Function that
|
|||
|
|
produces a list of candidates, given a certain knowledge base and a textual mention.
|
|||
|
|
get_candidates_batch (
|
|||
|
|
Callable[[KnowledgeBase, Iterable[Span]], Iterable[Iterable[Candidate]]], Iterable[Candidate]]
|
|||
|
|
): Function that produces a list of candidates, given a certain knowledge base and several textual mentions.
|
|||
|
|
generate_empty_kb (Callable[[Vocab, int], KnowledgeBase]): Callable returning empty KnowledgeBase.
|
|||
|
|
scorer (Optional[Callable]): The scoring method.
|
|||
|
|
use_gold_ents (bool): Whether to copy entities from gold docs during training or not. If false, another
|
|||
|
|
component must provide entity annotations.
|
|||
|
|
candidates_batch_size (int): Size of batches for entity candidate generation.
|
|||
|
|
threshold (Optional[float]): Confidence threshold for entity predictions. If confidence is below the threshold,
|
|||
|
|
prediction is discarded. If None, predictions are not filtered by any threshold.
|
|||
|
|
Úinclude_span_makerF)r( r) r* r+ r, r- r0 r1 )
|
|||
|
|
r( r) r* r+ r, r- r. r/ r0 r1 r2 r3 r4 )ÚattrsÚgetr" ÚvocabÚ EntityLinker)r< r= r% r( r) r* r+ r, r- r. r/ r0 r1 r2 r3 r4 s ú]C:\Users\garci\AppData\Roaming\Python\Python312\site-packages\spacy/pipeline/entity_linker.pyÚmake_entity_linkerrE + s� € ðN �;‰;�?‰?Ð/°Ô 7ä Ø �I‰IØ Ø Ø)ØØ!Ø%Ø!5Ø)Ø Øô
|
|||
|
|
ð
|
|||
|
|
ô Ø � ‰ Ø
|
|||
|
|
Ø Ø%ØØ Ø!Ø 1Ø%Ø 1Ø+ØØØ#Ø 3Øô! ð ó c óP — t j | fdt j gi|¤ŽS )NÚnegative_labels)r Ú score_linksrC ÚNIL)ÚexamplesÚkwargss rD Úentity_linker_scorerM • s& € Ü × Ñ ˜hÑ U¼ ×9IÑ9IÐ8JÐ UÈfÑ UÐUrF c ó — t S ©N)rM © rF rD Úmake_entity_linker_scorerrQ ™ s € ä Ð rF c $ óŽ — e Zd ZdZdZ d7eeddœdedede de
|
|||
|
|
e d e d
|
|||
|
|
e d e d e d
|
|||
|
|
e
|
|||
|
|
eege
|
|||
|
|
e f de
|
|||
|
|
ee
|
|||
|
|
e ge
|
|||
|
|
e
|
|||
|
|
e f de
|
|||
|
|
ee gef de dee
|
|||
|
|
de de dee ddf"d„Zde
|
|||
|
|
e de
|
|||
|
|
e fd„Zde
|
|||
|
|
egef fd„Zd8d„Zddd œd e
|
|||
|
|
g e
|
|||
|
|
e f d ee dee
|
|||
|
|
egef fd „Zd „ Zd!ddd"œde
|
|||
|
|
e d#ed$ee d%ee e ef de e ef f
|
|||
|
|
d&„Z de
|
|||
|
|
e d'e fd(„Z d)e
|
|||
|
|
e de!e fd*„Z"d)e
|
|||
|
|
e d+e!e ddfd,„Z# e$« d-œd.„Z% e$« d-œd/„Z& e'« d-œd0e(e e)f d1e
|
|||
|
|
e ddfd2„Z* e'« d-œd0e(e e)f d1e
|
|||
|
|
e dd fd3„Z+ddd4œd5„Z,d6„ Z-y)9rC z^Pipeline component for named entity linking.
|
|||
|
|
|
|||
|
|
DOCS: https://spacy.io/api/entitylinker
|
|||
|
|
rJ N)r0 r1 r4 rB r% r= r( r) r* r+ r, r- r. r/ r0 r1 r2 r3 r4 Úreturnc
|
|||
|
|
ó ‡ ‡
|
|||
|
|
— |�8d|cxk rdk s-n t t j j dd|¬« « ‚|‰ _ |‰ _ |‰ _ t |« ‰ _ |‰ _ |‰ _
|
|||
|
|
|‰ _ | ‰ _ |
|
|||
|
|
‰ _
|
|||
|
|
d| i‰ _ t d¬« ‰ _ | ‰ j |« ‰ _ |‰ _ |‰ _ |‰ _ |dk rt t j* « ‚dt, t. fˆ
|
|||
|
|
ˆ fd „ }|‰ _ y)
|
|||
|
|
aU Initialize an entity linker.
|
|||
|
|
|
|||
|
|
vocab (Vocab): The shared vocabulary.
|
|||
|
|
model (thinc.api.Model): The Thinc Model powering the pipeline component.
|
|||
|
|
name (str): The component instance name, used to add entries to the
|
|||
|
|
losses during training.
|
|||
|
|
labels_discard (Iterable[str]): NER labels that will automatically get a "NIL" prediction.
|
|||
|
|
n_sents (int): The number of neighbouring sentences to take into account.
|
|||
|
|
incl_prior (bool): Whether or not to include prior probabilities from the KB in the model.
|
|||
|
|
incl_context (bool): Whether or not to include the local context in the model.
|
|||
|
|
entity_vector_length (int): Size of encoding vectors in the KB.
|
|||
|
|
get_candidates (Callable[[KnowledgeBase, Span], Iterable[Candidate]]): Function that
|
|||
|
|
produces a list of candidates, given a certain knowledge base and a textual mention.
|
|||
|
|
get_candidates_batch (
|
|||
|
|
Callable[[KnowledgeBase, Iterable[Span]], Iterable[Iterable[Candidate]]],
|
|||
|
|
Iterable[Candidate]]
|
|||
|
|
): Function that produces a list of candidates, given a certain knowledge base and several textual mentions.
|
|||
|
|
generate_empty_kb (Callable[[Vocab, int], KnowledgeBase]): Callable returning empty KnowledgeBase.
|
|||
|
|
scorer (Optional[Callable]): The scoring method. Defaults to Scorer.score_links.
|
|||
|
|
use_gold_ents (bool): Whether to copy entities from gold docs or not. If false, another
|
|||
|
|
component must provide entity annotations.
|
|||
|
|
candidates_batch_size (int): Size of batches for entity candidate generation.
|
|||
|
|
threshold (Optional[float]): Confidence threshold for entity predictions. If confidence is below the
|
|||
|
|
threshold, prediction is discarded. If None, predictions are not filtered by any threshold.
|
|||
|
|
DOCS: https://spacy.io/api/entitylinker#init
|
|||
|
|
Nr r! )Ú range_startÚ range_endÚvaluer0 F)Ú normalizerK c óÐ •— ‰s‰S ‰j s ‰| fi |¤ŽS ‰j | « } ‰j d„ | D « « }t | |« D ] \ }}||_ Œ ‰| fi |¤ŽS )Nc 3 ó4 K — | ] }|j –— Œ y wrO )Ú predicted)Ú.0Úegs rD ú <genexpr>zFEntityLinker.__init__.<locals>._score_with_ents_set.<locals>.<genexpr>þ s è ø€ Ð5©H b�R—\•\©Hùs ‚)r2 Ú _ensure_entsÚpipeÚzipr[ )rK rL Údocsr] Údocr1 Úselfs €€rD Ú_score_with_ents_setz3EntityLinker.__init__.<locals>._score_with_ents_setô sz ø€ ñ Ø �
|
|||
|
|
Ø×%Ò%Ù ˜hÑ1¨&Ñ1Ð1à ×,Ñ,¨XÓ6�Ø—y‘yÙ5©HÓ5ó�ô # 8¨TÖ 2‘G�B˜Ø#&�B•Lð 3á ˜hÑ1¨&Ñ1Ð1rF )Ú
|
|||
|
|
ValueErrorr ÚE1043ÚformatrB r% r= Úlistr( r) r* r+ r- r. Úcfgr
|
|||
|
|
ÚdistanceÚkbr2 r3 r4 ÚE1044r r r1 )rd rB r% r= r( r) r* r+ r, r- r. r/ r0 r1 r2 r3 r4 re s ` ` rD Ú__init__zEntityLinker.__init__¦ s ù€ ðb Ð ¨!¨yÔ*=¸AÔ*=Ü Ü— ‘ ×#Ñ#Ø !Ø Ø#ð $ó óð
|
|||
|
|
𠈌
|
|||
|
|
؈Œ
|
|||
|
|
؈Œ Ü " >Ó 2ˆÔà ˆŒ Ø$ˆŒØ (ˆÔØ ,ˆÔØ$8ˆÔ!Ø$/°Ð#;ˆŒÜ&°Ô7ˆŒ
|
|||
|
|
Ù# D§J¡JÐ0DÓEˆŒØ *ˆÔØ%:ˆÔ"Ø"ˆŒà 1Ò $Ü œVŸ\™\Ó*Ð *ð 2¬8´GÑ+<ö 2ð +ˆ� rF rK c ó¾ — | j s|S g }|D ]G }|j « \ }}|j « }||j _ |j |« ŒI |S )zLIf use_gold_ents is true, set the gold entities to (a copy of) eg.predicted.)r2 Úget_aligned_ents_and_nerÚcopyr[ ÚentsÚappend)rd rK Ú new_examplesr] rr Ú_Únew_egs rD r_ zEntityLinker._ensure_ents s` € à×!Ò!؈Oàˆ ÛˆBØ×1Ñ1Ó3‰GˆD�!Ø—W‘W“YˆFØ$(ˆF× Ñ Ô !Ø × Ñ Õ 'ð ð
|
|||
|
|
ÐrF Ú kb_loaderc ó¬ — t |« s2t t j j t |« ¬« « ‚ || j « | _ y)ziDefine the KB of this pipe by providing a function that will
|
|||
|
|
create it using this object's vocab.)Úarg_typeN)Úcallablerf r ÚE885rh ÚtyperB rl )rd rw s rD Úset_kbzEntityLinker.set_kb s= € ô ˜ Ô"Ü œVŸ[™[× /Ñ /¼¸i» Ð /Ó IÓJÐ Já˜DŸJ™JÓ'ˆ�rF c óJ — | j €3t t j j | j
|
|||
|
|
¬« « ‚t
|
|||
|
|
| j d« rN| j j « r3t t j j | j
|
|||
|
|
¬« « ‚y y )N©r= Úis_empty) rl rf r ÚE1018rh r= Úhasattrr€ ÚE139©rd s rD Ú validate_kbzEntityLinker.validate_kb sr € à �7‰7ˆ?Ü œVŸ\™\× 0Ñ 0°d·i±iÐ 0Ó @ÓAÐ AÜ �4—7‘7˜JÔ '¨D¯G©G×,<Ñ,<Ô,>Ü œVŸ[™[× /Ñ /°T·Y±YÐ /Ó ?Ó@Ð @ð -?Ð 'rF )r< rw Ú get_examplesr< c ó˜ — t |d« |�| j |« | j « | j j }g }g }| j t
|
|||
|
|
|« d« « }|D ]S }|j } |j | « |j | j j j |« « ŒU t |« dkD s/J t j j | j ¬« « ‚t |« dkD s/J t j j | j ¬« « ‚t# |D � cg c] } | j$ ‘Œ c} « }
|
|||
|
|
|
|
|||
|
|
s|d } | dd } d| _ | f| _ | j j) || j j j+ |d¬ « ¬
|
|||
|
|
« |
|
|||
|
|
sg _ yyc c} w ) a’ Initialize the pipe for training, using a representative set
|
|||
|
|
of data examples.
|
|||
|
|
|
|||
|
|
get_examples (Callable[[], Iterable[Example]]): Function that
|
|||
|
|
returns a representative sample of gold-standard Example objects.
|
|||
|
|
nlp (Language): The current nlp object the component is part of.
|
|||
|
|
kb_loader (Callable[[Vocab], KnowledgeBase]): A function that creates a KnowledgeBase from a Vocab
|
|||
|
|
instance. Note that providing this argument will overwrite all data accumulated in the current KB.
|
|||
|
|
Use this only when loading a KB as-such from file.
|
|||
|
|
|
|||
|
|
DOCS: https://spacy.io/api/entitylinker#initialize
|
|||
|
|
zEntityLinker.initializeNé
|
|||
|
|
r r r! ÚXXXÚfloat32©Údtype)ÚXÚY)r r} r… rl r, r_ r Úxrs r% ÚopsÚalloc1fÚlenr ÚE923rh r= Úanyrr Úlabel_Ú
|
|||
|
|
initializeÚasarray) rd r† r< rw ÚnOÚ
|
|||
|
|
doc_sampleÚ
|
|||
|
|
vector_samplerK r] rc Úhas_annotationsÚents rD r– zEntityLinker.initialize" sŽ € ô& ˜lÐ,EÔFØ Ð Ø �K‰K˜ Ô "Ø ×ÑÔØ
|
|||
|
|
�W‰W×
|
|||
|
|
)Ñ
|
|||
|
|
)ˆØˆ
|
|||
|
|
؈
|
|||
|
|
Ø×$Ñ$¤V©L«N¸BÓ%?Ó@ˆÛˆBØ—$‘$ˆCØ × Ñ ˜cÔ "Ø × Ñ §¡§ ¡ ×!7Ñ!7¸Ó!;Õ <ð ô �:‹ Ò"ÐF¤F§K¡K×$6Ñ$6¸D¿I¹IÐ$6Ó$FÓFÐ"Ü�=Ó! AÒ%ÐI¤v§{¡{×'9Ñ'9¸t¿y¹yÐ'9Ó'IÓIÐ%ô ±:Ó >±:¨C˜sŸx›x°:Ñ >Ó?ˆÙ Ø ˜Q‘-ˆCØ�a˜�(ˆCØ ˆCŒJØ�vˆCŒHà �
|
|||
|
|
‰
|
|||
|
|
× Ñ Ø˜DŸJ™JŸN™N× 2Ñ 2°=È Ð 2Ó Rð ô
|
|||
|
|
ñ àˆC�Hð ùò ?s ÅGc óœ — |D ]G }|j j D ], }t | j | j |« « }|sŒ+ y ŒI y)z€Check if a batch contains a learnable example.
|
|||
|
|
|
|||
|
|
If one isn't present, then the update step needs to be skipped.
|
|||
|
|
TF)r[ rr ri r- rl )rd rK r] rœ Ú
|
|||
|
|
candidatess rD Úbatch_has_learnable_examplez(EntityLinker.batch_has_learnable_exampleX sH € ó ˆBØ—|‘|×(Ô(�Ü ! $×"5Ñ"5°d·g±g¸sÓ"CÓ D�
|
|||
|
|
Ú Ú ñ )ð ð rF ç )ÚdropÚsgdÚlossesr¡ r¢ r£ c ó — | j « |€i }|j | j d« |s|S | j |« }t |d« | j |« s|S t
|
|||
|
|
| j |« |D �cg c] }|j ‘Œ }}| j j |« \ }}| j ||¬« \ } }
|
|||
|
|
||
|
|||
|
|
« |�| j |« || j xx | z
|
|||
|
|
cc< |S c c}w )a. Learn from a batch of documents and gold-standard information,
|
|||
|
|
updating the pipe's model. Delegates to predict and get_loss.
|
|||
|
|
|
|||
|
|
examples (Iterable[Example]): A batch of Example objects.
|
|||
|
|
drop (float): The dropout rate.
|
|||
|
|
sgd (thinc.api.Optimizer): The optimizer.
|
|||
|
|
losses (Dict[str, float]): Optional record of the loss during training.
|
|||
|
|
Updated using the component name as the key.
|
|||
|
|
RETURNS (Dict[str, float]): The updated losses dictionary.
|
|||
|
|
|
|||
|
|
DOCS: https://spacy.io/api/entitylinker#update
|
|||
|
|
r zEntityLinker.update)Úsentence_encodingsrK ) r… Ú
|
|||
|
|
setdefaultr= r_ r rŸ r r% r[ Ú begin_updateÚget_lossÚ
|
|||
|
|
finish_update) rd rK r¡ r¢ r£ r] rb r¥ Ú
|
|||
|
|
bp_contextÚlossÚd_scoress rD ÚupdatezEntityLinker.updatef sý € ð(
|
|||
|
|
×ÑÔØ ˆ>؈FØ×ј$Ÿ)™) SÔ)Ù؈MØ×$Ñ$ XÓ.ˆÜ˜(Ð$9Ô:ð ×/Ñ/°Ô9؈M䘟™ TÔ*Ù'/Ó0¡x �— “ xˆÐ0Ø)-¯©×)@Ñ)@ÀÓ)FÑ&ИJà Ÿ ™ Ø 1¸Hð 'ó
|
|||
|
|
‰ˆˆhñ �8Ô Ø ˆ?Ø × Ñ ˜sÔ #؈t�y‰yÓ˜TÑ!Óàˆ
|
|||
|
|
ùò 1s Á?C<r¥ c ó| — t |d« g }d}g }|D ]} }|j dd¬« }|j « D ]U }||j } | r=| j j | « }
|
|||
|
|
|j
|
|||
|
|
|
|
|||
|
|
« |j
|
|||
|
|
|« |dz
|
|||
|
|
}ŒW Œ | j j j |d¬« }|| } |s1 | j j j |j Ž } d| fS | j |j k7 r,t j j d d
|
|||
|
|
¬ « }
|
|||
|
|
t |
|
|||
|
|
« ‚| j j# | |« } | j j j |j Ž } || |< | j j% | |« }|t' |« z }t) |« | fS ) NzEntityLinker.get_lossr Ú ENT_KB_IDT)Ú as_stringr! rŠ r‹ r¨ z gold entities do not match up©ÚmethodÚmsg)r Ú get_alignedÚget_matching_entsÚstartrl Ú
|
|||
|
|
get_vectorrs r% r� Ú asarray2fÚalloc2fÚshaper ÚE147rh Ú RuntimeErrorrk Úget_gradr¨ r’ Úfloat)rd rK r¥ Úentity_encodingsÚeidxÚ keep_entsr] Úkb_idsrœ Úkb_idÚentity_encodingÚselected_encodingsÚoutÚerrÚ gradientsr« s rD r¨ zEntityLinker.get_loss• s« € ܘ(Ð$;Ô<Ø ÐàˆØˆ ãˆBØ—^‘^ K¸4�^Ó@ˆFà×+Ñ+Ö-�Ø ˜sŸy™yÑ)�ÙØ&*§g¡g×&8Ñ&8¸Ó&?�OØ$×+Ñ+¨OÔ<Ø ×$Ñ$ TÔ*à˜‘ ‘ñ .ð ð Ÿ:™:Ÿ>™>×3Ñ3Ð4DÈIÐ3ÓVÐØ /° Ñ :Ðñ Ø(�$—*‘*—.‘.×(Ñ(Ð*<×*BÑ*BÐCˆCØ�c�6ˆMà × #Ñ #Ð'7×'=Ñ'=Ò =Ü—+‘+×$Ñ$Ø!Ð'Fð %ó ˆCô ˜sÓ#Ð #Ø—M‘M×*Ñ*Ð+=Ð?OÓPˆ à$ˆd�j‰j�n‰n×$Ñ$Ð&8×&>Ñ&>Ð?ˆØ"ˆˆI‰à�}‰}×%Ñ%Ð&8Ð:JÓKˆØ”cÐ*Ó+Ñ+ˆÜ�T‹{˜CÐ Ð rF rb c ó — | j « d}g }| j j j }|s|S t |t
|
|||
|
|
« r|g}t
|
|||
|
|
|« D �]Í \ }}t |« dk( rŒ|j D �cg c] }|‘Œ }}t dt |j « | j « D �]q } |j | | | j z }
|
|||
|
|
t t |
|
|||
|
|
« « D � cg c] } |
|
|||
|
|
| j | j vr| ‘Œ! } } t | j dkD r-| j | j | D � cg c] } |
|
|||
|
|
| ‘Œ c} « n,| D � cg c]! } | j# | j |
|
|||
|
|
| « ‘Œ# c} « }
|
|||
|
|
t
|
|||
|
|
|
|
|||
|
|
« D �]• \ }}t% |d« sJ ‚t |j « }|j' |d « |j' |d « f}|d |d cxk\ rdk\ sJ ‚ J ‚| j( r»t+ d|d | j, z
|
|||
|
|
« }t/ t |« dz
|
|||
|
|
|d | j, z « }|| j0 }|| j2 }||| j5 « }| j j7 |g« d }|j8 }|j: j= |« }|dz
|
|||
|
|
}|j | j v r |j? | j@ « �Œit |
|
|||
|
|
| « }|s |j? | j@ « �Œ–t |« dk( r,| jB € |j? |d jD « �ŒÐtG jH |« |jK |D �cg c] }|jL ‘Œ c}« } | jN s |jK |D � cg c] } d‘Œ c} « } | } | j( rÜ|jK |D �cg c] }|jP ‘Œ c}« } |j: j= | d¬« } t | « t | « k7 r*tS tT jV jY dd¬ « « ‚|j[ | « | z z }!|!j\ | j\ k7 rt_ tT j` « ‚| |!z | |!z z
|
|||
|
|
} |j? | jB � | j+ « | jB k\ r+|| jc « je « jD ntf j@ « �Œ˜ �Œt �ŒÐ t |« |k( s,tT jV jY dd
|
|||
|
|
¬ « }"tS |"« ‚|S c c}w c c} w c c} w c c} w c c}w c c} w c c}w ) ap Apply the pipeline's model to a batch of docs, without modifying them.
|
|||
|
|
Returns the KB IDs for each entity in each doc, including NIL if there is
|
|||
|
|
no prediction.
|
|||
|
|
|
|||
|
|
docs (Iterable[Doc]): The documents to predict.
|
|||
|
|
RETURNS (List[str]): The models prediction for each document.
|
|||
|
|
|
|||
|
|
DOCS: https://spacy.io/api/entitylinker#predict
|
|||
|
|
r r! Úsentséÿÿÿÿr )ÚaxisÚpredictzvectors not of equal lengthr± z$result variables not of equal length)4r… r% r� ÚxpÚ
|
|||
|
|
isinstancer Ú enumerater’ rÊ Úrangerr r3 r• r( ri r. rl r- r‚ Úindexr+ Úmaxr) Úminr¶ ÚendÚas_docrÍ ÚTÚlinalgÚnormrs rJ r4 Úentity_ÚrandomÚshuffler— Ú
|
|||
|
|
prior_probr* Ú
|
|||
|
|
entity_vectorr¼ r r» rh Údotrº rf ÚE161ÚargmaxÚitemrC )#rd rb Ú entity_countÚ final_kb_idsrÎ Úirc ÚsÚ sentencesÚent_idxÚ ent_batchÚidxÚ
|
|||
|
|
valid_ent_idxÚbatch_candidatesÚjrœ rÊ Ú sent_indicesÚstart_sentenceÚ end_sentenceÚ start_tokenÚ end_tokenÚsent_docÚsentence_encodingÚsentence_encoding_tÚ
|
|||
|
|
sentence_normrž ÚcÚ prior_probsru Úscoresr¿ Ú entity_normÚsimsrÇ s# rD rÍ zEntityLinker.predict½ s# € ð
|
|||
|
|
×ÑÔ؈ Ø"$ˆ Ø
|
|||
|
|
�Z‰Z�^‰^×
|
|||
|
|
Ñ
|
|||
|
|
ˆÙØ Ð Ü �dœCÔ Ø�6ˆDÜ —o‰FˆAˆsÜ�3‹x˜1Š}ØØ$'§I¢IÓ.¡I˜qš IˆIÐ.ô ! ¤C¨¯©£M°4×3MÑ3M×N�Ø ŸH™H W¨w¸×9SÑ9SÑ/SÐ T� ô
|
|||
|
|
%¤S¨£^Ô 4ó!á 4˜Ø ‘~×,Ñ,°D×4GÑ4GÑGò Ø 4ð ð !ô $(ð ×1Ñ1°AÒ5ð ×-Ñ-Ø Ÿ™¹MÓ!J¹M°S )¨C£.¸MÑ!Jôñ $1óá#0˜Cð ×+Ñ+¨D¯G©G°Y¸s±^ÕDØ#0ñó $Ð ô (¨ × 2‘F�A�sÜ" 3¨Ô0Ð0Ð0Ü §¡›O�Eà!Ÿ™¨¨a©Ó1Ø!Ÿ™¨¨b© Ó2ð$�Lð (¨™?¨l¸1©oÔBÀÒBÐBÑBÐBÐBà×(Ò(ä),¨Q° ¸Q±À$Ç,Á,Ñ0NÓ)O˜Ü'*Ü ›N¨QÑ .° ¸Q±À$Ç,Á,Ñ0Nó(˜ ð '0°Ñ&?×&EÑ&E˜ Ø$-¨lÑ$;×$?Ñ$?˜ Ø#& {°9Ð#=×#DÑ#DÓ#F˜ð -1¯J©J×,>Ñ,>À¸zÓ,JÈ1Ñ,MÐ)Ø.?×.AÑ.AÐ+Ø(*¯ © ¯©Ð7JÓ(K˜
|
|||
|
|
Ø AÑ%�LØ—z‘z T×%8Ñ%8Ñ8à$×+Ñ+¨D¯H©HÖ5ä%)Ð*:¸1Ñ*=Ó%>˜
|
|||
|
|
Ù )à (× /Ñ /°·±Ö 9Ü ›_°Ò 1°d·n±nÐ6Là (× /Ñ /°
|
|||
|
|
¸1±
|
|||
|
|
×0EÑ0EÖ Fä "ŸN™N¨:Ô 6à*,¯*©*ÉJÓ5WÉJÀq°a·l³lÈJÑ5WÓ*X˜KØ#'§?¢?Ø.0¯j©jÁzÓ9RÁzÀ!º#ÀzÑ9RÓ.S Ø%0˜Fà #× 0Ò 0Ø35·:±:Ù>HÓ$I¹j¸ Q§_£_¸jÑ$Ió4"Ð 0ð /1¯i©i¯n©nÐ=MÐTU¨nÓ.V Ü#&Ð'7Ó#8¼CÀ Ó<LÒ#LÜ*6Ü(.¯ © ×(:Ñ(:Ø3<Ø0Mð );ó )*ó+&ð %&ð (*§v¡vÐ.>Ð@SÓ'TØ$1°KÑ$?ñ(" ð $(§:¡:°×1BÑ1BÒ#BÜ*4´V·[±[Ó*AÐ$AØ)4°tÑ);¸{ÈTÑ?QÑ)R Ø (× /Ñ /à#'§>¡>Ð#9Ø#)§:¡:£<°4·>±>Ò#Að !+¨6¯=©=«?×+?Ñ+?Ó+AÑ B× JÒ Jô &2×%5Ñ%5ö ò} 3ò- Oð
|
|||
|
|
&ôB �LÓ! \Ò1Ü—+‘+×$Ñ$Ø Ð&Lð %ó ˆCô ˜sÓ#Ð #ØÐùòG /ùò!ùò "KùòùòZ 6Xùâ9Rùò
|
|||
|
|
%Js* Á; U$Ã&$U)Ä: U.Å&U3 Í=U8Î2 U=Ï Vr c ó� — t |D ��cg c] }|j D ] }|‘Œ Œ c}}« }|t |« k7 r3t t j j |t |« ¬« « ‚d}| j d }|D ]= }|j D ], }|| }|dz
|
|||
|
|
}|D ] } | j dk( s|sŒ|| _ Œ Œ. Œ? yc c}}w )a Modify a batch of documents, using pre-computed scores.
|
|||
|
|
|
|||
|
|
docs (Iterable[Doc]): The documents to modify.
|
|||
|
|
kb_ids (List[str]): The IDs to set, produced by EntityLinker.predict.
|
|||
|
|
|
|||
|
|
DOCS: https://spacy.io/api/entitylinker#set_annotations
|
|||
|
|
)rr Úidsr r0 r! N) r’ rr rf r ÚE148rh rj Ú ent_kb_idÚ
|
|||
|
|
ent_kb_id_)
|
|||
|
|
rd rb r rc rœ Ú
|
|||
|
|
count_entsrå r0 rà Útokens
|
|||
|
|
rD Úset_annotationsz EntityLinker.set_annotations7 s» € ô ©ÔB© #¸¿¼°#š#¸˜#¨ÒBÓCˆ
|