course/pr_1021/rum/_app/immutable/nodes/8.da90ff27.js

import{s as O,n as I,o as K}from"../chunks/scheduler.37c15a92.js";import{S as V,i as W,g as l,s as n,r as H,A as X,h as s,f as a,c as r,j as G,u as B,x as b,k as J,y as Z,a as i,v as j,d as S,t as U,w as q}from"../chunks/index.2bf4358c.js";import{Y as ee}from"../chunks/Youtube.1e50a667.js";import{C as te}from"../chunks/CourseFloatingBanner.6add7356.js";import{H as ae,E as ie}from"../chunks/getInferenceSnippets.ebf8be91.js";function ne(N){let o,x,_,E,m,w,c,C,u,z,p,Q="Modelele Encoder utilizează doar encoderul unui model Transformer. La fiecare etapă, layer-urile de atenție pot accesa toate cuvintele din propoziția inițială. Aceste modele sunt adesea caracterizate ca având o atenție „bidirecțională” și sunt adesea numite <em>modele auto-encoding</em>.",P,f,Y="Preantrenarea acestor modele se bazează, de obicei, pe alterarea unei propoziții date (de exemplu, prin mascarea unor cuvinte aleatorii) și pe sarcina modelului de a găsi sau reconstrui propoziția inițială.",T,d,k="Modelele Encoder sunt cele mai potrivite pentru sarcinile care necesită înțelegerea întregii propoziții, cum ar fi clasificarea propozițiilor, recunoașterea entităților numite (și, mai general, clasificarea cuvintelor) și Extractive QA.",M,$,D="Printre reprezentanții acestei familii de modele se numără:",L,g,F='<li><a href="https://huggingface.co/docs/transformers/model_doc/albert" rel="nofollow">ALBERT</a></li> <li><a href="https://huggingface.co/docs/transformers/model_doc/bert" rel="nofollow">BERT</a></li> <li><a href="https://huggingface.co/docs/transformers/model_doc/distilbert" rel="nofollow">DistilBERT</a></li> <li><a href="https://huggingface.co/docs/transformers/model_doc/electra" rel="nofollow">ELECTRA</a></li> <li><a href="https://huggingface.co/docs/transformers/model_doc/roberta" rel="nofollow">RoBERTa</a></li>',R,h,y,v,A;return m=new ae({props:{title:"Modele Encoder",local:"modele-encoder",headingTag:"h1"}}),c=new te({props:{chapter:1,classNames:"absolute z-10 right-0 top-0"}}),u=new ee({props:{id:"MUqNwgPjJvQ"}}),h=new ie({props:{source:"https://github.com/huggingface/course/blob/main/chapters/rum/chapter1/5.mdx"}}),{c(){o=l("meta"),x=n(),_=l("p"),E=n(),H(m.$$.fragment),w=n(),H(c.$$.fragment),C=n(),H(u.$$.fragment),z=n(),p=l("p"),p.innerHTML=Q,P=n(),f=l("p"),f.textContent=Y,T=n(),d=l("p"),d.textContent=k,M=n(),$=l("p"),$.textContent=D,L=n(),g=l("ul"),g.innerHTML=F,R=n(),H(h.$$.fragment),y=n(),v=l("p"),this.h()},l(e){const t=X("svelte-u9bgzb",document.head);o=s(t,"META",{name:!0,content:!0}),t.forEach(a),x=r(e),_=s(e,"P",{}),G(_).forEach(a),E=r(e),B(m.$$.fragment,e),w=r(e),B(c.$$.fragment,e),C=r(e),B(u.$$.fragment,e),z=r(e),p=s(e,"P",{"data-svelte-h":!0}),b(p)!=="svelte-7cte6f"&&(p.innerHTML=Q),P=r(e),f=s(e,"P",{"data-svelte-h":!0}),b(f)!=="svelte-1ube0l3"&&(f.textContent=Y),T=r(e),d=s(e,"P",{"data-svelte-h":!0}),b(d)!=="svelte-jpa42"&&(d.textContent=k),M=r(e),$=s(e,"P",{"data-svelte-h":!0}),b($)!=="svelte-dd7odx"&&($.textContent=D),L=r(e),g=s(e,"UL",{"data-svelte-h":!0}),b(g)!=="svelte-17o0nd4"&&(g.innerHTML=F),R=r(e),B(h.$$.fragment,e),y=r(e),v=s(e,"P",{}),G(v).forEach(a),this.h()},h(){J(o,"name","hf:doc:metadata"),J(o,"content",re)},m(e,t){Z(document.head,o),i(e,x,t),i(e,_,t),i(e,E,t),j(m,e,t),i(e,w,t),j(c,e,t),i(e,C,t),j(u,e,t),i(e,z,t),i(e,p,t),i(e,P,t),i(e,f,t),i(e,T,t),i(e,d,t),i(e,M,t),i(e,$,t),i(e,L,t),i(e,g,t),i(e,R,t),j(h,e,t),i(e,y,t),i(e,v,t),A=!0},p:I,i(e){A||(S(m.$$.fragment,e),S(c.$$.fragment,e),S(u.$$.fragment,e),S(h.$$.fragment,e),A=!0)},o(e){U(m.$$.fragment,e),U(c.$$.fragment,e),U(u.$$.fragment,e),U(h.$$.fragment,e),A=!1},d(e){e&&(a(x),a(_),a(E),a(w),a(C),a(z),a(p),a(P),a(f),a(T),a(d),a(M),a($),a(L),a(g),a(R),a(y),a(v)),a(o),q(m,e),q(c,e),q(u,e),q(h,e)}}}const re='{"title":"Modele Encoder","local":"modele-encoder","sections":[],"depth":1}';function oe(N){return K(()=>{new URLSearchParams(window.location.search).get("fw")}),[]}class pe extends V{constructor(o){super(),W(this,o,oe,ne,O,{})}}export{pe as component};