course/pr_1202/en/_app/immutable/nodes/77.dbac4dcd.js

import{s as X,o as Y}from"../chunks/scheduler.893fe8c9.js";import{S as Z,i as ee,e as f,s as c,c as M,h as te,a as m,d as n,b as h,f as V,g as x,j as g,k as D,l as ne,m as r,n as P,t as H,o as E,p as z}from"../chunks/index.b1df2166.js";import{C as ae,H as ie,E as re}from"../chunks/MermaidChart.svelte_svelte_type_style_lang.4edfee50.js";import{C as oe}from"../chunks/CourseFloatingBanner.c1c08878.js";import{F as le}from"../chunks/FrameworkSwitchCourse.4480e339.js";function se(y){let a,p='To do this, you’ll need to leverage everything you learned about training models with the Keras API in <a href="/course/chapter3">Chapter 3</a>, the 🤗 Datasets library in <a href="/course/chapter5">Chapter 5</a>, and the 🤗 Tokenizers library in <a href="/course/chapter6">Chapter 6</a>. We’ll also upload our results to the Model Hub, like we did in <a href="/course/chapter4">Chapter 4</a>, so this is really the chapter where everything comes together!',l,o,u="Each section can be read independently.";return{c(){a=f("p"),a.innerHTML=p,l=c(),o=f("p"),o.textContent=u},l(i){a=m(i,"P",{"data-svelte-h":!0}),g(a)!=="svelte-1203ezd"&&(a.innerHTML=p),l=h(i),o=m(i,"P",{"data-svelte-h":!0}),g(o)!=="svelte-1q4m2vs"&&(o.textContent=u)},m(i,s){r(i,a,s),r(i,l,s),r(i,o,s)},d(i){i&&(n(a),n(l),n(o))}}}function ue(y){let a,p='To do this, you’ll need to leverage everything you learned about the <code>Trainer</code> API and the 🤗 Accelerate library in <a href="/course/chapter3">Chapter 3</a>, the 🤗 Datasets library in <a href="/course/chapter5">Chapter 5</a>, and the 🤗 Tokenizers library in <a href="/course/chapter6">Chapter 6</a>. We’ll also upload our results to the Model Hub, like we did in <a href="/course/chapter4">Chapter 4</a>, so this is really the chapter where everything comes together!',l,o,u="Each section can be read independently and will show you how to train a model with the <code>Trainer</code> API or with your own training loop, using 🤗 Accelerate. Feel free to skip either part and focus on the one that interests you the most: the <code>Trainer</code> API is great for fine-tuning or training your model without worrying about what’s going on behind the scenes, while the training loop with <code>Accelerate</code> will let you customize any part you want more easily.";return{c(){a=f("p"),a.innerHTML=p,l=c(),o=f("p"),o.innerHTML=u},l(i){a=m(i,"P",{"data-svelte-h":!0}),g(a)!=="svelte-j9oz82"&&(a.innerHTML=p),l=h(i),o=m(i,"P",{"data-svelte-h":!0}),g(o)!=="svelte-gf723x"&&(o.innerHTML=u)},m(i,s){r(i,a,s),r(i,l,s),r(i,o,s)},d(i){i&&(n(a),n(l),n(o))}}}function ce(y){let a,p,l,o,u,i,s,I,$,A,b,q,_,G='In <a href="/course/chapter3">Chapter 3</a>, you saw how to fine-tune a model for text classification. In this chapter, we will tackle the following common language tasks that are essential for working with both traditional NLP models and modern LLMs:',S,v,K="<li>Token classification</li> <li>Masked language modeling (like BERT)</li> <li>Summarization</li> <li>Translation</li> <li>Causal language modeling pretraining (like GPT-2)</li> <li>Question answering</li>",F,k,Q="These fundamental tasks form the foundation of how Large Language Models (LLMs) work and understanding them is crucial for effectively working with today’s most advanced language models.",B,T,w,R="<p>If you read the sections in sequence, you will notice that they have quite a bit of code and prose in common. The repetition is intentional, to allow you to dip in (or come back later) to any task that interests you and find a complete working example.</p>",N,C,O,L,U;u=new le({props:{fw:y[0]}}),s=new ae({props:{containerStyle:"float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"}}),$=new ie({props:{title:"Introduction",local:"introduction",headingTag:"h1"}}),b=new oe({props:{chapter:7,classNames:"absolute z-10 right-0 top-0"}});function W(e,t){return e[0]==="pt"?ue:se}let j=W(y),d=j(y);return C=new re({props:{source:"https://github.com/huggingface/course/blob/main/chapters/en/chapter7/1.mdx"}}),{c(){a=f("meta"),p=c(),l=f("p"),o=c(),M(u.$$.fragment),i=c(),M(s.$$.fragment),I=c(),M($.$$.fragment),A=c(),M(b.$$.fragment),q=c(),_=f("p"),_.innerHTML=G,S=c(),v=f("ul"),v.innerHTML=K,F=c(),k=f("p"),k.textContent=Q,B=c(),d.c(),T=c(),w=f("blockquote"),w.innerHTML=R,N=c(),M(C.$$.fragment),O=c(),L=f("p"),this.h()},l(e){const t=te("svelte-u9bgzb",document.head);a=m(t,"META",{name:!0,content:!0}),t.forEach(n),p=h(e),l=m(e,"P",{}),V(l).forEach(n),o=h(e),x(u.$$.fragment,e),i=h(e),x(s.$$.fragment,e),I=h(e),x($.$$.fragment,e),A=h(e),x(b.$$.fragment,e),q=h(e),_=m(e,"P",{"data-svelte-h":!0}),g(_)!=="svelte-95s2kh"&&(_.innerHTML=G),S=h(e),v=m(e,"UL",{"data-svelte-h":!0}),g(v)!=="svelte-146pccd"&&(v.innerHTML=K),F=h(e),k=m(e,"P",{"data-svelte-h":!0}),g(k)!=="svelte-pq4i9y"&&(k.textContent=Q),B=h(e),d.l(e),T=h(e),w=m(e,"BLOCKQUOTE",{class:!0,"data-svelte-h":!0}),g(w)!=="svelte-v8ani2"&&(w.innerHTML=R),N=h(e),x(C.$$.fragment,e),O=h(e),L=m(e,"P",{}),V(L).forEach(n),this.h()},h(){D(a,"name","hf:doc:metadata"),D(a,"content",he),D(w,"class","tip")},m(e,t){ne(document.head,a),r(e,p,t),r(e,l,t),r(e,o,t),P(u,e,t),r(e,i,t),P(s,e,t),r(e,I,t),P($,e,t),r(e,A,t),P(b,e,t),r(e,q,t),r(e,_,t),r(e,S,t),r(e,v,t),r(e,F,t),r(e,k,t),r(e,B,t),d.m(e,t),r(e,T,t),r(e,w,t),r(e,N,t),P(C,e,t),r(e,O,t),r(e,L,t),U=!0},p(e,[t]){const J={};t&1&&(J.fw=e[0]),u.$set(J),j!==(j=W(e))&&(d.d(1),d=j(e),d&&(d.c(),d.m(T.parentNode,T)))},i(e){U||(H(u.$$.fragment,e),H(s.$$.fragment,e),H($.$$.fragment,e),H(b.$$.fragment,e),H(C.$$.fragment,e),U=!0)},o(e){E(u.$$.fragment,e),E(s.$$.fragment,e),E($.$$.fragment,e),E(b.$$.fragment,e),E(C.$$.fragment,e),U=!1},d(e){e&&(n(p),n(l),n(o),n(i),n(I),n(A),n(q),n(_),n(S),n(v),n(F),n(k),n(B),n(T),n(w),n(N),n(O),n(L)),n(a),z(u,e),z(s,e),z($,e),z(b,e),d.d(e),z(C,e)}}}const he='{"title":"Introduction","local":"introduction","sections":[],"depth":1}';function pe(y,a,p){let l="pt";return Y(()=>{const o=new URLSearchParams(window.location.search);p(0,l=o.get("fw")||"pt")}),[l]}class ye extends Z{constructor(a){super(),ee(this,a,pe,ce,X,{})}}export{ye as component};