INTUIA/Programa final/spacy/pipeline/__pycache__/tok2vec.cpython-312.pyc

Ë
>û g�5ã	ób
—dd
lm
Z

ddlmZmZmZmZmZmZm	Z	
ddl
mZmZm
Z
mZ
ddlmZ
ddlmZ
ddlmZ
ddlmZmZmZ
dd	lmZ
d
dlmZ
dZe«j=e«
d
Zej@ddg
d
ei
¬«dede!d
eddfd„«Z"Gd„de«Z#Gd„de«Z$d
e$de%fd„Z&d„Z'y)é)
Úislice)ÚAnyÚCallableÚDictÚIterableÚListÚOptionalÚSequence)ÚConfigÚModelÚ	OptimizerÚset_dropout_rateé)
ÚErrors)
ÚLanguage)
ÚDoc)ÚExampleÚvalidate_examplesÚvalidate_get_examples)
ÚVocabé
)
Ú
TrainablePipez­
[model]
@architectures = "spacy.HashEmbedCNN.v2"
pretrained_vectors = null
width = 96
depth = 4
embed_size = 2000
window_size = 1
maxout_pieces = 3
subword_features = true
ÚmodelÚtok2vecz
doc.tensor)ÚassignsÚdefault_configÚnlpÚnameÚreturnÚTok2Veccó0—t
|j||
«S©
N)r Úvocab)rrrs   úWC:\Users\garci\AppData\Roaming\Python\Python312\site-packages\spacy/pipeline/tok2vec.pyÚmake_tok2vecr%s€ô�3—9‘9˜e TÓ*Ð*ócó2
—eZ
dZd
Zd dedededdfd„Zede	dfd	„«Z
ede	efd
„«Zdddeddfd
„Zdddede
fd„Zd!d„Zdeefd„Zdeeddfd„Zddddœdeededeedeeeeffd„Zd!d„Zddœ
degeefdeefd„Zd„Zy)"r aAApply a "token-to-vector" model and set its outputs in the doc.tensor
    attribute. This is mostly useful to share a single subnetwork between multiple
    components, e.g. to have one embedding and CNN network shared between a
    parser, tagger and NER.

    In order to use the `Tok2Vec` predictions, subsequent components should use
    the `Tok2VecListener` layer as the tok2vec subnetwork of their model. This
    layer will read data from the `doc.tensor` attribute during prediction.
    During training, the `Tok2Vec` component will save its prediction and backprop
    callback for each batch, so that the subsequent components can backpropagate
    to the shared weights. This implementation is used because it allows us to
    avoid relying on object identity within the models to achieve the parameter
    sharing.
    r#rrrNcóJ—|
|_||_
||_i|_i|_y
)a“
Initialize a tok2vec component.

        vocab (Vocab): The shared vocabulary.
        model (thinc.api.Model[List[Doc], List[Floats2d]]):
            The Thinc Model powering the pipeline component. It should take
            a list of Doc objects as input, and output a list of 2d float arrays.
        name (str): The component instance name.

        DOCS: https://spacy.io/api/tok2vec#init
        N)r#rrÚlistener_mapÚcfg)Úselfr#rrs    r$Ú__init__zTok2Vec.__init__2s(€ðˆŒ
؈Œ
؈Œ	Ø@BˆÔØ#%ˆ�r&ÚTok2VecListenerc
ón—|jD�
�cgc]}
|j|
D]}|‘ŒŒc}}
Sc
c}}
w)
zuRETURNS (List[Tok2VecListener]): The listener models listening to this
        component. Usually internals.
        )Úlistening_componentsr))r+Ú
cÚ
ms   r$Ú	listenerszTok2Vec.listenersCs9€ð
 ×4Ò4ÔSÑ4�a¸d×>OÑ>OÐPQÔ>R¸’
Ð>R�
Ð4ÒSÐSùÓSs�1c
óH—t
|jj««
S)
zoRETURNS (List[str]): The downstream components listening to this
        component. Usually internals.
        )Úlistr)Úkeys)
r+s
 r$r/zTok2Vec.listening_componentsJs€ô
�D×%Ñ%×*Ñ*Ó,Ó-Ð-r&ÚlistenerÚcomponent_namecóœ—|jj|g«
|
|j|v
r|j|j|
«

y
y
)z=Add a listener for a downstream component. Usually internals.N)r)Ú
setdefaultÚappend©r+r6r7s   r$Úadd_listenerzTok2Vec.add_listenerQsI€à×Ñ×$Ñ$ ^°RÔ8ؘ4×,Ñ,¨^Ñ<Ñ<Ø×Ñ˜nÑ-×4Ñ4°XÕ>ð=r&có¸—||jvrL|
|j|vr;|j|j|
«

|j|s
|j|=y
y)z@Remove a listener for a downstream component. Usually internals.TF)r)Úremover;s   r$Úremove_listenerzTok2Vec.remove_listenerWs^€à˜T×.Ñ.Ñ.ؘ4×,Ñ,¨^Ñ<Ñ<Ø×!Ñ! .Ñ1×8Ñ8¸ÔBà×(Ñ(¨Ò8Ø×)Ñ)¨.Ð9ØØr&có
—d
|jf}tt|
dd«t«r\|
jj«D]>}t|t«s
Œ|j|vs
Œ#|j||
j«
Œ@yy)a¬
Walk over a model of a processing component, looking for layers that
        are Tok2vecListener subclasses that have an upstream_name that matches
        this component. Listeners can also set their upstream_name attribute to
        the wildcard string '*' to match any `Tok2Vec`.

        You're unlikely to ever need multiple `Tok2Vec` components, so it's
        fine to leave your listeners upstream_name on '*'.
        Ú
*rN)	rÚ
isinstanceÚgetattrrrÚwalkr-Ú
upstream_namer<)r+Ú	componentÚnamesÚnodes    r$Úfind_listenerszTok2Vec.find_listenersbsm€ð�d—i‘iÐ ˆÜ”g˜i¨°$Ó7¼Ô?Ø!Ÿ™×,Ñ,Ö.�ܘd¤OÕ4¸×9KÑ9KÈuÒ9TØ×%Ñ% d¨I¯N©NÕ;ñ/ð@
r&Údocscó

—t
d
„|
D««
sP|jjd«
}|
D�cgc])}|jjj	d|f«
‘Œ+c}S|jj|
«
}|Sc
c}w)a?
Apply the pipeline's model to a batch of docs, without modifying them.
        Returns a single tensor for a batch of documents.

        docs (Iterable[Doc]): The documents to predict.
        RETURNS: Vector representations for each token in the documents.

        DOCS: https://spacy.io/api/tok2vec#predict
        c
3ó2K
—|]}
t
|
«
–
—

Œy­w
r")
Úlen©Ú.0Údocs  r$ú	<genexpr>z"Tok2Vec.predict.<locals>.<genexpr>zsèø€Ð,¡t ”3�s—8¡tùs‚
ÚnOr)ÚanyrÚget_dimÚopsÚallocÚpredict)r+rJÚwidthrPÚtokvecss     r$rWzTok2Vec.predictqsq€ôÑ,¡tÓ,Ô,à—J‘J×&Ñ& tÓ,ˆEÙ@DÓEÁ¸�D—J‘J—N‘N×(Ñ(¨!¨U¨Õ4ÀÑEÐEØ—*‘*×$Ñ$ TÓ*ˆØˆùòF
s².Bcót—t
|
|«D])\}}|jd
t|«
k(sJ‚
||_Œ+y)zøModify a batch of documents, using pre-computed scores.

        docs (Iterable[Doc]): The documents to modify.
        tokvecses: The tensors to set, produced by Tok2Vec.predict.

        DOCS: https://spacy.io/api/tok2vec#set_annotations
        rN)ÚzipÚshaperMÚtensor)r+rJÚ	tokvecsesrPrYs     r$Úset_annotationszTok2Vec.set_annotations�s;€ô   iÖ0‰LˆC�Ø—=‘= Ñ#¤s¨3£xÒ/Ð/Ð/Ø ˆC�Jñ1r&ç)ÚdropÚsgdÚlossesÚexamplesrarbrcc󂇇‡‡‡‡
‡—‰€iŠt
|
d«
|
D�cgc]}|j‘Œ}}t‰j|«
‰jj	|«
\ŠŠ‰D�cgc]/}‰jj
j|jŽ‘Œ1c}Š
‰j‰jd«
ˆ
ˆˆˆfd„Šˆˆˆ
ˆˆfd„}tj|«
}	‰jd
dD]}
|
j|	‰‰«
Œ‰jr ‰jdj|	‰|«
‰Sc
c}wc
c}w)aLearn from a batch of documents and gold-standard information,
        updating the pipe's model.

        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/tok2vec#update
        NzTok2Vec.updater`c
ó>
•—t
t|«
«
D]F}
‰|
xx||
z
cc<‰‰jxxt||
d
zj	««
z
cc<ŒH‰D�cgc]/}‰j
jj|jŽ‘Œ1c}Sc
c}w)zžAccumulate tok2vec loss and gradient. This is passed as a callback
            to all but the last listener. Only the last one does the backprop.
            r)	ÚrangerMrÚfloatÚsumrrUÚalloc2fr\)Ú
one_d_tokvecsÚ
iÚt2vÚ	d_tokvecsrcr+rYs   €€€€r$Úaccumulate_gradientz+Tok2Vec.update.<locals>.accumulate_gradientªsŽø€ô
œ3˜}Ó-Ö.�
ؘ!“ 
¨aÑ 0Ñ0“Ø�t—y‘yÓ!¤U¨M¸!Ñ,<À
Ñ,A×+FÑ+FÓ+HÓ%IÑIÔ!ð/ñC
J
ÓJÁ'¸3Ð*�D—J‘J—N‘N×*Ñ*¨C¯I©IÒ6À'ÑJÐJùÒJsÁ#4Bc
óN•—‰|«

‰‰«
}
‰�‰j
‰«

|
S)
z>Callback to actually do the backprop. Passed to last listener.)
Ú
finish_update)rkÚd_docsroÚ
bp_tokvecsrnr+rbs  €€€€€r$Úbackpropz Tok2Vec.update.<locals>.backprop´s-ø€á 
Ô.Ù 	Ó*ˆF؈Ø×"Ñ" 3Ô'؈Mr&éÿÿÿÿ)rÚ	predictedrrÚbegin_updaterUrjr\r9rr-Úget_batch_idr2Úreceive)r+rdrarbrcÚegrJrmrtÚbatch_idr6rorsrnrYs`  ``      @@@@r$ÚupdatezTok2Vec.update�s
þ€ð(ˆ>؈Fܘ(Ð$4Ô5Ù'/Ó0¡x �—“ xˆÐ0ܘŸ™ TÔ*Ø"Ÿj™j×5Ñ5°dÓ;ш�ÙCJÓKÁ7¸CÐ+�T—Z‘Z—^‘^×+Ñ+¨S¯Y©YÒ7À7ÑKˆ	Ø×ј$Ÿ)™) SÔ)÷	K
÷	ð	ô#×/Ñ/°Ó5ˆØŸ™ s¨Ó+ˆHØ×Ñ˜X wÐ0CÕDð,à�>Š>Ø�N‰N˜2Ñ×&Ñ& x°¸(ÔC؈
ùò;1ùòL
sœD7Á)4D<có—yr"©)r+rdÚscoress   r$Úget_losszTok2Vec.get_lossÃs€Ør&)
rÚget_examplesrc
ó
—t
|
d
«
g}t|
«d«D]}|j|j«

Œ|s/Jtj
j
|j¬«
«‚
|jj|¬«

y)at
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.

        DOCS: https://spacy.io/api/tok2vec#initialize
        zTok2Vec.initializeé
©
r)
Ú
XN)
rrr:Ú
xrÚE923ÚformatrrÚ
initialize)r+r�rÚ
doc_sampleÚexamples     r$r‰zTok2Vec.initializeÆsq€ô	˜lÐ,@ÔA؈
Ü™l›n¨bÖ1ˆGØ×Ñ˜gŸi™iÕ(ð2áÐ=œ6Ÿ;™;×-Ñ-°4·9±9Ð-Ó=Ó=ˆzØ�
‰
×Ñ 
ÐÕ+r&c
ó—t‚
r")
ÚNotImplementedError)r+Úlabels  r$Ú	add_labelzTok2Vec.add_labelÜs€Ü!Ð!r&)
r)rN) Ú__name__Ú
__module__Ú__qualname__Ú__doc__rrÚstrr,Úpropertyrr2r/r<Úboolr?rIrrrWr
r_rrhr	r
rr|r€rrr‰r�r~r&r$r r "sX
„ñ
ñ&˜eð&¨Eð&¸ð&ÈTó&ð"ðT
˜4Ð 1Ñ2òT
óðT
ðð. d¨3¡iò.óð.ð?Ð%6ð?Èð?ÐPTó?ð	Ð(9ð	È3ð	ÐSWó	ó
<ð˜H S™Móð 
! H¨S¡Mð
!Àó
!ð Ø#'Ø-1ò
4à˜7Ñ#ð4ðð	4ð
�iÑ
 ð4ð˜˜c 5˜jÑ)Ñ*ó
4ól


ð#'ò	,à˜r 8¨GÑ#4Ð4Ñ5ð,ð�hÑ
ó	,ó,
"r&cód—eZ
dZd
ZdZdededdfd„Zede	e
defd	„«Zd
eddfd„Zde