course/pr_1101/en/_app/immutable/nodes/4.f92f2ea2.js

import{s as Ge,n as Ye,o as Fe}from"../chunks/scheduler.893fe8c9.js";import{S as Ke,i as We,e as s,s as l,c as m,h as Oe,a as r,d as i,b as a,f as De,g as p,j as o,k as je,l as Je,m as n,n as d,t as f,o as u,p as c}from"../chunks/index.b1df2166.js";import{C as Qe,H as $,E as Ve}from"../chunks/MermaidChart.svelte_svelte_type_style_lang.812bcf2e.js";import{C as Xe}from"../chunks/CourseFloatingBanner.0a6f233d.js";function Ze(Me){let h,W,F,O,g,J,v,Q,L,V,T,Ce="In this chapter, you’ve been introduced to the fundamentals of Transformer models, Large Language Models (LLMs), and how they’re revolutionizing AI and beyond.",X,w,Z,x,ee,y,be="We explored what NLP is and how Large Language Models have transformed the field. You learned that:",te,M,_e="<li>NLP encompasses a wide range of tasks from classification to generation</li> <li>LLMs are powerful models trained on massive amounts of text data</li> <li>These models can perform multiple tasks within a single architecture</li> <li>Despite their capabilities, LLMs have limitations including hallucinations and bias</li>",ie,C,ne,b,Pe="You saw how the <code>pipeline()</code> function from 🤗 Transformers makes it easy to use pre-trained models for various tasks:",le,_,He="<li>Text classification, token classification, and question answering</li> <li>Text generation and summarization</li> <li>Translation and other sequence-to-sequence tasks</li> <li>Speech recognition and image classification</li>",ae,P,se,H,ke="We discussed how Transformer models work at a high level, including:",re,k,Ee="<li>The importance of the attention mechanism</li> <li>How transfer learning enables models to adapt to specific tasks</li> <li>The three main architectural variants: encoder-only, decoder-only, and encoder-decoder</li>",oe,E,me,q,qe="A key aspect of this chapter was understanding which architecture to use for different tasks:",pe,z,ze="<thead><tr><th>Model</th> <th>Examples</th> <th>Tasks</th></tr></thead> <tbody><tr><td>Encoder-only</td> <td>BERT, DistilBERT, ModernBERT</td> <td>Sentence classification, named entity recognition, extractive question answering</td></tr> <tr><td>Decoder-only</td> <td>GPT, LLaMA, Gemma, SmolLM</td> <td>Text generation, conversational AI, creative writing</td></tr> <tr><td>Encoder-decoder</td> <td>BART, T5, Marian, mBART</td> <td>Summarization, translation, generative question answering</td></tr></tbody>",de,S,fe,A,Se="You also learned about recent developments in the field:",ue,U,Ae="<li>How LLMs have grown in size and capability over time</li> <li>The concept of scaling laws and how they guide model development</li> <li>Specialized attention mechanisms that help models process longer sequences</li> <li>The two-phase training approach of pretraining and instruction tuning</li>",ce,B,he,I,Ue="Throughout the chapter, you’ve seen how these models can be applied to real-world problems:",$e,N,Be="<li>Using the Hugging Face Hub to find and use pre-trained models</li> <li>Leveraging the Inference API to test models directly in your browser</li> <li>Understanding which models are best suited for specific tasks</li>",ge,R,ve,D,Ie="Now that you have a solid understanding of what Transformer models are and how they work at a high level, you’re ready to dive deeper into how to use them effectively. In the next chapters, you’ll learn how to:",Le,j,Ne="<li>Use the Transformers library to load and fine-tune models</li> <li>Process different types of data for model input</li> <li>Adapt pre-trained models to your specific tasks</li> <li>Deploy models for practical applications</li>",Te,G,Re="The foundation you’ve built in this chapter will serve you well as you explore more advanced topics and techniques in the coming sections.",we,Y,xe,K,ye;return g=new Qe({props:{containerStyle:"float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"}}),v=new $({props:{title:"Summary",local:"summary",headingTag:"h1"}}),L=new Xe({props:{chapter:1,classNames:"absolute z-10 right-0 top-0"}}),w=new $({props:{title:"Key concepts covered",local:"key-concepts-covered",headingTag:"h2"}}),x=new $({props:{title:"Natural Language Processing and LLMs",local:"natural-language-processing-and-llms",headingTag:"h3"}}),C=new $({props:{title:"Transformer capabilities",local:"transformer-capabilities",headingTag:"h3"}}),P=new $({props:{title:"Transformer architecture",local:"transformer-architecture",headingTag:"h3"}}),E=new $({props:{title:"Model architectures and their applications",local:"model-architectures-and-their-applications",headingTag:"h3"}}),S=new $({props:{title:"Modern LLM developments",local:"modern-llm-developments",headingTag:"h3"}}),B=new $({props:{title:"Practical applications",local:"practical-applications",headingTag:"h3"}}),R=new $({props:{title:"Looking ahead",local:"looking-ahead",headingTag:"h2"}}),Y=new Ve({props:{source:"https://github.com/huggingface/course/blob/main/chapters/en/chapter1/10.mdx"}}),{c(){h=s("meta"),W=l(),F=s("p"),O=l(),m(g.$$.fragment),J=l(),m(v.$$.fragment),Q=l(),m(L.$$.fragment),V=l(),T=s("p"),T.textContent=Ce,X=l(),m(w.$$.fragment),Z=l(),m(x.$$.fragment),ee=l(),y=s("p"),y.textContent=be,te=l(),M=s("ul"),M.innerHTML=_e,ie=l(),m(C.$$.fragment),ne=l(),b=s("p"),b.innerHTML=Pe,le=l(),_=s("ul"),_.innerHTML=He,ae=l(),m(P.$$.fragment),se=l(),H=s("p"),H.textContent=ke,re=l(),k=s("ul"),k.innerHTML=Ee,oe=l(),m(E.$$.fragment),me=l(),q=s("p"),q.textContent=qe,pe=l(),z=s("table"),z.innerHTML=ze,de=l(),m(S.$$.fragment),fe=l(),A=s("p"),A.textContent=Se,ue=l(),U=s("ul"),U.innerHTML=Ae,ce=l(),m(B.$$.fragment),he=l(),I=s("p"),I.textContent=Ue,$e=l(),N=s("ul"),N.innerHTML=Be,ge=l(),m(R.$$.fragment),ve=l(),D=s("p"),D.textContent=Ie,Le=l(),j=s("ul"),j.innerHTML=Ne,Te=l(),G=s("p"),G.textContent=Re,we=l(),m(Y.$$.fragment),xe=l(),K=s("p"),this.h()},l(e){const t=Oe("svelte-u9bgzb",document.head);h=r(t,"META",{name:!0,content:!0}),t.forEach(i),W=a(e),F=r(e,"P",{}),De(F).forEach(i),O=a(e),p(g.$$.fragment,e),J=a(e),p(v.$$.fragment,e),Q=a(e),p(L.$$.fragment,e),V=a(e),T=r(e,"P",{"data-svelte-h":!0}),o(T)!=="svelte-i3fshz"&&(T.textContent=Ce),X=a(e),p(w.$$.fragment,e),Z=a(e),p(x.$$.fragment,e),ee=a(e),y=r(e,"P",{"data-svelte-h":!0}),o(y)!=="svelte-1lit2zm"&&(y.textContent=be),te=a(e),M=r(e,"UL",{"data-svelte-h":!0}),o(M)!=="svelte-17sn8p4"&&(M.innerHTML=_e),ie=a(e),p(C.$$.fragment,e),ne=a(e),b=r(e,"P",{"data-svelte-h":!0}),o(b)!=="svelte-jhecjx"&&(b.innerHTML=Pe),le=a(e),_=r(e,"UL",{"data-svelte-h":!0}),o(_)!=="svelte-1shrou"&&(_.innerHTML=He),ae=a(e),p(P.$$.fragment,e),se=a(e),H=r(e,"P",{"data-svelte-h":!0}),o(H)!=="svelte-1f8v9li"&&(H.textContent=ke),re=a(e),k=r(e,"UL",{"data-svelte-h":!0}),o(k)!=="svelte-1q9ne17"&&(k.innerHTML=Ee),oe=a(e),p(E.$$.fragment,e),me=a(e),q=r(e,"P",{"data-svelte-h":!0}),o(q)!=="svelte-1ftf9m3"&&(q.textContent=qe),pe=a(e),z=r(e,"TABLE",{"data-svelte-h":!0}),o(z)!=="svelte-19df94o"&&(z.innerHTML=ze),de=a(e),p(S.$$.fragment,e),fe=a(e),A=r(e,"P",{"data-svelte-h":!0}),o(A)!=="svelte-t6urrr"&&(A.textContent=Se),ue=a(e),U=r(e,"UL",{"data-svelte-h":!0}),o(U)!=="svelte-1re5xby"&&(U.innerHTML=Ae),ce=a(e),p(B.$$.fragment,e),he=a(e),I=r(e,"P",{"data-svelte-h":!0}),o(I)!=="svelte-1uuqwsg"&&(I.textContent=Ue),$e=a(e),N=r(e,"UL",{"data-svelte-h":!0}),o(N)!=="svelte-ha9wcz"&&(N.innerHTML=Be),ge=a(e),p(R.$$.fragment,e),ve=a(e),D=r(e,"P",{"data-svelte-h":!0}),o(D)!=="svelte-9vvrc8"&&(D.textContent=Ie),Le=a(e),j=r(e,"UL",{"data-svelte-h":!0}),o(j)!=="svelte-lytq9t"&&(j.innerHTML=Ne),Te=a(e),G=r(e,"P",{"data-svelte-h":!0}),o(G)!=="svelte-9onihf"&&(G.textContent=Re),we=a(e),p(Y.$$.fragment,e),xe=a(e),K=r(e,"P",{}),De(K).forEach(i),this.h()},h(){je(h,"name","hf:doc:metadata"),je(h,"content",et)},m(e,t){Je(document.head,h),n(e,W,t),n(e,F,t),n(e,O,t),d(g,e,t),n(e,J,t),d(v,e,t),n(e,Q,t),d(L,e,t),n(e,V,t),n(e,T,t),n(e,X,t),d(w,e,t),n(e,Z,t),d(x,e,t),n(e,ee,t),n(e,y,t),n(e,te,t),n(e,M,t),n(e,ie,t),d(C,e,t),n(e,ne,t),n(e,b,t),n(e,le,t),n(e,_,t),n(e,ae,t),d(P,e,t),n(e,se,t),n(e,H,t),n(e,re,t),n(e,k,t),n(e,oe,t),d(E,e,t),n(e,me,t),n(e,q,t),n(e,pe,t),n(e,z,t),n(e,de,t),d(S,e,t),n(e,fe,t),n(e,A,t),n(e,ue,t),n(e,U,t),n(e,ce,t),d(B,e,t),n(e,he,t),n(e,I,t),n(e,$e,t),n(e,N,t),n(e,ge,t),d(R,e,t),n(e,ve,t),n(e,D,t),n(e,Le,t),n(e,j,t),n(e,Te,t),n(e,G,t),n(e,we,t),d(Y,e,t),n(e,xe,t),n(e,K,t),ye=!0},p:Ye,i(e){ye||(f(g.$$.fragment,e),f(v.$$.fragment,e),f(L.$$.fragment,e),f(w.$$.fragment,e),f(x.$$.fragment,e),f(C.$$.fragment,e),f(P.$$.fragment,e),f(E.$$.fragment,e),f(S.$$.fragment,e),f(B.$$.fragment,e),f(R.$$.fragment,e),f(Y.$$.fragment,e),ye=!0)},o(e){u(g.$$.fragment,e),u(v.$$.fragment,e),u(L.$$.fragment,e),u(w.$$.fragment,e),u(x.$$.fragment,e),u(C.$$.fragment,e),u(P.$$.fragment,e),u(E.$$.fragment,e),u(S.$$.fragment,e),u(B.$$.fragment,e),u(R.$$.fragment,e),u(Y.$$.fragment,e),ye=!1},d(e){e&&(i(W),i(F),i(O),i(J),i(Q),i(V),i(T),i(X),i(Z),i(ee),i(y),i(te),i(M),i(ie),i(ne),i(b),i(le),i(_),i(ae),i(se),i(H),i(re),i(k),i(oe),i(me),i(q),i(pe),i(z),i(de),i(fe),i(A),i(ue),i(U),i(ce),i(he),i(I),i($e),i(N),i(ge),i(ve),i(D),i(Le),i(j),i(Te),i(G),i(we),i(xe),i(K)),i(h),c(g,e),c(v,e),c(L,e),c(w,e),c(x,e),c(C,e),c(P,e),c(E,e),c(S,e),c(B,e),c(R,e),c(Y,e)}}}const et='{"title":"Summary","local":"summary","sections":[{"title":"Key concepts covered","local":"key-concepts-covered","sections":[{"title":"Natural Language Processing and LLMs","local":"natural-language-processing-and-llms","sections":[],"depth":3},{"title":"Transformer capabilities","local":"transformer-capabilities","sections":[],"depth":3},{"title":"Transformer architecture","local":"transformer-architecture","sections":[],"depth":3},{"title":"Model architectures and their applications","local":"model-architectures-and-their-applications","sections":[],"depth":3},{"title":"Modern LLM developments","local":"modern-llm-developments","sections":[],"depth":3},{"title":"Practical applications","local":"practical-applications","sections":[],"depth":3}],"depth":2},{"title":"Looking ahead","local":"looking-ahead","sections":[],"depth":2}],"depth":1}';function tt(Me){return Fe(()=>{new URLSearchParams(window.location.search).get("fw")}),[]}class st extends Ke{constructor(h){super(),We(this,h,tt,Ze,Ge,{})}}export{st as component};