Buckets:

hf-doc-build
/

doc-dev

hf-doc-build/doc-dev / transformers /main /ko /_app /immutable /nodes /20.86a089fc.js

rtrm's picture

about 2 months ago

21 kB

	import{s as ks,n as As,o as Qs}from"../chunks/scheduler.9bc65507.js";import{S as Rs,i as xs,g as p,s as e,r as i,A as Gs,h as M,f as l,c as n,j as Ys,u as y,x as r,k as Vs,y as Ns,a as t,v as J,d as m,t as c,w as T}from"../chunks/index.707bf1b6.js";import{C as U}from"../chunks/CodeBlock.54a9f38d.js";import{H as us,E as Hs}from"../chunks/EditOnGithub.922df6ba.js";function Ss(Is){let o,S,N,z,j,F,h,ds="🤗 Transformers에서는 🤗 Transformers 모델을 학습시키는데 최적화된 <code>Trainer</code> 클래스를 제공하기 때문에, 사용자는 직접 훈련 루프를 작성할 필요 없이 더욱 간편하게 학습을 시킬 수 있습니다. 또한, <code>Trainer</code>는 하이퍼파라미터 탐색을 위한 API를 제공합니다. 이 문서에서 이 API를 활용하는 방법을 예시와 함께 보여드리겠습니다.",E,w,D,u,Cs=`<code>Trainer</code>는 현재 아래 4가지 하이퍼파라미터 탐색 백엔드를 지원합니다:
	<a href="https://optuna.org/" rel="nofollow">optuna</a>와 <a href="https://sigopt.com/" rel="nofollow">sigopt</a>, <a href="https://docs.ray.io/en/latest/tune/index.html" rel="nofollow">raytune</a>, <a href="https://wandb.ai/site/sweeps" rel="nofollow">wandb</a> 입니다.`,L,I,fs="하이퍼파라미터 탐색 백엔드로 사용하기 전에 아래의 명령어를 사용하여 라이브러리들을 설치하세요.",P,d,K,C,O,f,bs="하이퍼파라미터 탐색 공간을 정의하세요. 하이퍼파라미터 탐색 백엔드마다 서로 다른 형식이 필요합니다.",ss,b,gs='sigopt의 경우, 해당 <a href="https://docs.sigopt.com/ai-module-api-references/api_reference/objects/object_parameter" rel="nofollow">object_parameter</a> 문서를 참조하여 아래와 같이 작성하세요:',as,g,ls,_,_s='optuna의 경우, 해당 <a href="https://optuna.readthedocs.io/en/stable/tutorial/10_key_features/002_configurations.html#sphx-glr-tutorial-10-key-features-002-configurations-py" rel="nofollow">object_parameter</a> 문서를 참조하여 아래와 같이 작성하세요:',ts,$,es,q,$s='raytune의 경우, 해당 <a href="https://docs.ray.io/en/latest/tune/api/search_space.html" rel="nofollow">object_parameter</a> 문서를 참조하여 아래와 같이 작성하세요:',ns,W,ps,X,qs='wandb의 경우, 해당 <a href="https://docs.wandb.ai/guides/sweeps/configuration" rel="nofollow">object_parameter</a> 문서를 참조하여 아래와 같이 작성하세요:',Ms,Z,is,B,Ws="<code>model_init</code> 함수를 정의하고 이를 <code>Trainer</code>에 전달하세요. 아래는 그 예시입니다.",ys,v,rs,Y,Xs="아래와 같이 <code>model_init</code> 함수, 훈련 인수, 훈련 및 테스트 데이터셋, 그리고 평가 함수를 사용하여 <code>Trainer</code>를 생성하세요:",Js,V,ms,k,Zs="하이퍼파라미터 탐색을 호출하고, 최적의 시험 매개변수를 가져오세요. 백엔드는 <code>"optuna"</code>/<code>"sigopt"</code>/<code>"wandb"</code>/<code>"ray"</code> 중에서 선택할 수 있습니다. 방향은 <code>"minimize"</code> 또는 <code>"maximize"</code> 중 선택하며, 목표를 최소화할 것인지 최대화할 것인지를 결정합니다.",cs,A,Bs="자신만의 compute_objective 함수를 정의할 수 있습니다. 만약 이 함수를 정의하지 않으면, 기본 compute_objective가 호출되고, f1과 같은 평가 지표의 합이 목푯값으로 반환됩니다.",Ts,Q,os,R,Us,x,vs="현재, DDP(Distributed Data Parallelism; 분산 데이터 병렬처리)를 위한 하이퍼파라미터 탐색은 optuna와 sigopt에서 가능합니다. 최상위 프로세스가 하이퍼파라미터 탐색 과정을 시작하고 그 결과를 다른 프로세스에 전달합니다.",js,G,hs,H,ws;return j=new us({props:{title:"Trainer API를 사용한 하이퍼파라미터 탐색",local:"hyperparameter-search-using-trainer-api",headingTag:"h1"}}),w=new us({props:{title:"하이퍼파라미터 탐색 백엔드",local:"hyperparameter-search-backend",headingTag:"h2"}}),d=new U({props:{code:"cGlwJTIwaW5zdGFsbCUyMG9wdHVuYSUyRnNpZ29wdCUyRndhbmRiJTJGcmF5JTVCdHVuZSU1RCUyMA==",highlighted:"pip install optuna/sigopt/wandb/ray[tune] ",wrap:!1}}),C=new us({props:{title:"예제에서 하이퍼파라미터 탐색을 활성화하는 방법",local:"how-to-enable-hyperparameter-search-in-example",headingTag:"h2"}}),g=new U({props:{code:"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",highlighted:`<span class="hljs-meta">>>> </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">sigopt_hp_space</span>(<span class="hljs-params">trial</span>):
	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> [
	<span class="hljs-meta">... </span> {<span class="hljs-string">"bounds"</span>: {<span class="hljs-string">"min"</span>: <span class="hljs-number">1e-6</span>, <span class="hljs-string">"max"</span>: <span class="hljs-number">1e-4</span>}, <span class="hljs-string">"name"</span>: <span class="hljs-string">"learning_rate"</span>, <span class="hljs-string">"type"</span>: <span class="hljs-string">"double"</span>},
	<span class="hljs-meta">... </span> {
	<span class="hljs-meta">... </span> <span class="hljs-string">"categorical_values"</span>: [<span class="hljs-string">"16"</span>, <span class="hljs-string">"32"</span>, <span class="hljs-string">"64"</span>, <span class="hljs-string">"128"</span>],
	<span class="hljs-meta">... </span> <span class="hljs-string">"name"</span>: <span class="hljs-string">"per_device_train_batch_size"</span>,
	<span class="hljs-meta">... </span> <span class="hljs-string">"type"</span>: <span class="hljs-string">"categorical"</span>,
	<span class="hljs-meta">... </span> },
	<span class="hljs-meta">... </span> ]`,wrap:!1}}),$=new U({props:{code:"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",highlighted:`<span class="hljs-meta">>>> </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">optuna_hp_space</span>(<span class="hljs-params">trial</span>):
	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> {
	<span class="hljs-meta">... </span> <span class="hljs-string">"learning_rate"</span>: trial.suggest_float(<span class="hljs-string">"learning_rate"</span>, <span class="hljs-number">1e-6</span>, <span class="hljs-number">1e-4</span>, log=<span class="hljs-literal">True</span>),
	<span class="hljs-meta">... </span> <span class="hljs-string">"per_device_train_batch_size"</span>: trial.suggest_categorical(<span class="hljs-string">"per_device_train_batch_size"</span>, [<span class="hljs-number">16</span>, <span class="hljs-number">32</span>, <span class="hljs-number">64</span>, <span class="hljs-number">128</span>]),
	<span class="hljs-meta">... </span> }`,wrap:!1}}),W=new U({props:{code:"ZGVmJTIwcmF5X2hwX3NwYWNlKHRyaWFsKSUzQSUwQSUyMCUyMCUyMCUyMHJldHVybiUyMCU3QiUwQSUyMCUyMCUyMCUyMCUyMCUyMCUyMCUyMCUyMmxlYXJuaW5nX3JhdGUlMjIlM0ElMjB0dW5lLmxvZ3VuaWZvcm0oMWUtNiUyQyUyMDFlLTQpJTJDJTBBJTIwJTIwJTIwJTIwJTIwJTIwJTIwJTIwJTIycGVyX2RldmljZV90cmFpbl9iYXRjaF9zaXplJTIyJTNBJTIwdHVuZS5jaG9pY2UoJTVCMTYlMkMlMjAzMiUyQyUyMDY0JTJDJTIwMTI4JTVEKSUyQyUwQSUyMCUyMCUyMCUyMCU3RA==",highlighted:`<span class="hljs-meta">>>> </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">ray_hp_space</span>(<span class="hljs-params">trial</span>):
	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> {
	<span class="hljs-meta">... </span> <span class="hljs-string">"learning_rate"</span>: tune.loguniform(<span class="hljs-number">1e-6</span>, <span class="hljs-number">1e-4</span>),
	<span class="hljs-meta">... </span> <span class="hljs-string">"per_device_train_batch_size"</span>: tune.choice([<span class="hljs-number">16</span>, <span class="hljs-number">32</span>, <span class="hljs-number">64</span>, <span class="hljs-number">128</span>]),
	<span class="hljs-meta">... </span> }`,wrap:!1}}),Z=new U({props:{code:"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",highlighted:`<span class="hljs-meta">>>> </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">wandb_hp_space</span>(<span class="hljs-params">trial</span>):
	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> {
	<span class="hljs-meta">... </span> <span class="hljs-string">"method"</span>: <span class="hljs-string">"random"</span>,
	<span class="hljs-meta">... </span> <span class="hljs-string">"metric"</span>: {<span class="hljs-string">"name"</span>: <span class="hljs-string">"objective"</span>, <span class="hljs-string">"goal"</span>: <span class="hljs-string">"minimize"</span>},
	<span class="hljs-meta">... </span> <span class="hljs-string">"parameters"</span>: {
	<span class="hljs-meta">... </span> <span class="hljs-string">"learning_rate"</span>: {<span class="hljs-string">"distribution"</span>: <span class="hljs-string">"uniform"</span>, <span class="hljs-string">"min"</span>: <span class="hljs-number">1e-6</span>, <span class="hljs-string">"max"</span>: <span class="hljs-number">1e-4</span>},
	<span class="hljs-meta">... </span> <span class="hljs-string">"per_device_train_batch_size"</span>: {<span class="hljs-string">"values"</span>: [<span class="hljs-number">16</span>, <span class="hljs-number">32</span>, <span class="hljs-number">64</span>, <span class="hljs-number">128</span>]},
	<span class="hljs-meta">... </span> },
	<span class="hljs-meta">... </span> }`,wrap:!1}}),v=new U({props:{code:"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",highlighted:`<span class="hljs-meta">>>> </span><span class="hljs-keyword">def</span> <span class="hljs-title function_">model_init</span>(<span class="hljs-params">trial</span>):
	<span class="hljs-meta">... </span> <span class="hljs-keyword">return</span> AutoModelForSequenceClassification.from_pretrained(
	<span class="hljs-meta">... </span> model_args.model_name_or_path,
	<span class="hljs-meta">... </span> from_tf=<span class="hljs-built_in">bool</span>(<span class="hljs-string">".ckpt"</span> <span class="hljs-keyword">in</span> model_args.model_name_or_path),
	<span class="hljs-meta">... </span> config=config,
	<span class="hljs-meta">... </span> cache_dir=model_args.cache_dir,
	<span class="hljs-meta">... </span> revision=model_args.model_revision,
	<span class="hljs-meta">... </span> token=<span class="hljs-literal">True</span> <span class="hljs-keyword">if</span> model_args.use_auth_token <span class="hljs-keyword">else</span> <span class="hljs-literal">None</span>,
	<span class="hljs-meta">... </span> )`,wrap:!1}}),V=new U({props:{code:"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",highlighted:`<span class="hljs-meta">>>> </span>trainer = Trainer(
	<span class="hljs-meta">... </span> model=<span class="hljs-literal">None</span>,
	<span class="hljs-meta">... </span> args=training_args,
	<span class="hljs-meta">... </span> train_dataset=small_train_dataset,
	<span class="hljs-meta">... </span> eval_dataset=small_eval_dataset,
	<span class="hljs-meta">... </span> compute_metrics=compute_metrics,
	<span class="hljs-meta">... </span> tokenizer=tokenizer,
	<span class="hljs-meta">... </span> model_init=model_init,
	<span class="hljs-meta">... </span> data_collator=data_collator,
	<span class="hljs-meta">... </span>)`,wrap:!1}}),Q=new U({props:{code:"YmVzdF90cmlhbCUyMCUzRCUyMHRyYWluZXIuaHlwZXJwYXJhbWV0ZXJfc2VhcmNoKCUwQSUyMCUyMCUyMCUyMGRpcmVjdGlvbiUzRCUyMm1heGltaXplJTIyJTJDJTBBJTIwJTIwJTIwJTIwYmFja2VuZCUzRCUyMm9wdHVuYSUyMiUyQyUwQSUyMCUyMCUyMCUyMGhwX3NwYWNlJTNEb3B0dW5hX2hwX3NwYWNlJTJDJTBBJTIwJTIwJTIwJTIwbl90cmlhbHMlM0QyMCUyQyUwQSUyMCUyMCUyMCUyMGNvbXB1dGVfb2JqZWN0aXZlJTNEY29tcHV0ZV9vYmplY3RpdmUlMkMlMEEp",highlighted:`<span class="hljs-meta">>>> </span>best_trial = trainer.hyperparameter_search(
	<span class="hljs-meta">... </span> direction=<span class="hljs-string">"maximize"</span>,
	<span class="hljs-meta">... </span> backend=<span class="hljs-string">"optuna"</span>,
	<span class="hljs-meta">... </span> hp_space=optuna_hp_space,
	<span class="hljs-meta">... </span> n_trials=<span class="hljs-number">20</span>,
	<span class="hljs-meta">... </span> compute_objective=compute_objective,
	<span class="hljs-meta">... </span>)`,wrap:!1}}),R=new us({props:{title:"DDP 미세 조정을 위한 하이퍼파라미터 탐색",local:"hyperparameter-search-for-ddp-finetune",headingTag:"h2"}}),G=new Hs({props:{source:"https://github.com/huggingface/transformers/blob/main/docs/source/ko/hpo_train.md"}}),{c(){o=p("meta"),S=e(),N=p("p"),z=e(),i(j.$$.fragment),F=e(),h=p("p"),h.innerHTML=ds,E=e(),i(w.$$.fragment),D=e(),u=p("p"),u.innerHTML=Cs,L=e(),I=p("p"),I.textContent=fs,P=e(),i(d.$$.fragment),K=e(),i(C.$$.fragment),O=e(),f=p("p"),f.textContent=bs,ss=e(),b=p("p"),b.innerHTML=gs,as=e(),i(g.$$.fragment),ls=e(),_=p("p"),_.innerHTML=_s,ts=e(),i($.$$.fragment),es=e(),q=p("p"),q.innerHTML=$s,ns=e(),i(W.$$.fragment),ps=e(),X=p("p"),X.innerHTML=qs,Ms=e(),i(Z.$$.fragment),is=e(),B=p("p"),B.innerHTML=Ws,ys=e(),i(v.$$.fragment),rs=e(),Y=p("p"),Y.innerHTML=Xs,Js=e(),i(V.$$.fragment),ms=e(),k=p("p"),k.innerHTML=Zs,cs=e(),A=p("p"),A.textContent=Bs,Ts=e(),i(Q.$$.fragment),os=e(),i(R.$$.fragment),Us=e(),x=p("p"),x.textContent=vs,js=e(),i(G.$$.fragment),hs=e(),H=p("p"),this.h()},l(s){const a=Gs("svelte-u9bgzb",document.head);o=M(a,"META",{name:!0,content:!0}),a.forEach(l),S=n(s),N=M(s,"P",{}),Ys(N).forEach(l),z=n(s),y(j.$$.fragment,s),F=n(s),h=M(s,"P",{"data-svelte-h":!0}),r(h)!=="svelte-1wpadea"&&(h.innerHTML=ds),E=n(s),y(w.$$.fragment,s),D=n(s),u=M(s,"P",{"data-svelte-h":!0}),r(u)!=="svelte-14yr5bp"&&(u.innerHTML=Cs),L=n(s),I=M(s,"P",{"data-svelte-h":!0}),r(I)!=="svelte-1n3obee"&&(I.textContent=fs),P=n(s),y(d.$$.fragment,s),K=n(s),y(C.$$.fragment,s),O=n(s),f=M(s,"P",{"data-svelte-h":!0}),r(f)!=="svelte-1aiezuq"&&(f.textContent=bs),ss=n(s),b=M(s,"P",{"data-svelte-h":!0}),r(b)!=="svelte-wifdo4"&&(b.innerHTML=gs),as=n(s),y(g.$$.fragment,s),ls=n(s),_=M(s,"P",{"data-svelte-h":!0}),r(_)!=="svelte-15tcz6x"&&(_.innerHTML=_s),ts=n(s),y($.$$.fragment,s),es=n(s),q=M(s,"P",{"data-svelte-h":!0}),r(q)!=="svelte-1q8ydvq"&&(q.innerHTML=$s),ns=n(s),y(W.$$.fragment,s),ps=n(s),X=M(s,"P",{"data-svelte-h":!0}),r(X)!=="svelte-1a9peeq"&&(X.innerHTML=qs),Ms=n(s),y(Z.$$.fragment,s),is=n(s),B=M(s,"P",{"data-svelte-h":!0}),r(B)!=="svelte-1q7kvtf"&&(B.innerHTML=Ws),ys=n(s),y(v.$$.fragment,s),rs=n(s),Y=M(s,"P",{"data-svelte-h":!0}),r(Y)!=="svelte-17wxp1q"&&(Y.innerHTML=Xs),Js=n(s),y(V.$$.fragment,s),ms=n(s),k=M(s,"P",{"data-svelte-h":!0}),r(k)!=="svelte-ikq9mj"&&(k.innerHTML=Zs),cs=n(s),A=M(s,"P",{"data-svelte-h":!0}),r(A)!=="svelte-1k38imx"&&(A.textContent=Bs),Ts=n(s),y(Q.$$.fragment,s),os=n(s),y(R.$$.fragment,s),Us=n(s),x=M(s,"P",{"data-svelte-h":!0}),r(x)!=="svelte-13bo63v"&&(x.textContent=vs),js=n(s),y(G.$$.fragment,s),hs=n(s),H=M(s,"P",{}),Ys(H).forEach(l),this.h()},h(){Vs(o,"name","hf:doc:metadata"),Vs(o,"content",zs)},m(s,a){Ns(document.head,o),t(s,S,a),t(s,N,a),t(s,z,a),J(j,s,a),t(s,F,a),t(s,h,a),t(s,E,a),J(w,s,a),t(s,D,a),t(s,u,a),t(s,L,a),t(s,I,a),t(s,P,a),J(d,s,a),t(s,K,a),J(C,s,a),t(s,O,a),t(s,f,a),t(s,ss,a),t(s,b,a),t(s,as,a),J(g,s,a),t(s,ls,a),t(s,_,a),t(s,ts,a),J($,s,a),t(s,es,a),t(s,q,a),t(s,ns,a),J(W,s,a),t(s,ps,a),t(s,X,a),t(s,Ms,a),J(Z,s,a),t(s,is,a),t(s,B,a),t(s,ys,a),J(v,s,a),t(s,rs,a),t(s,Y,a),t(s,Js,a),J(V,s,a),t(s,ms,a),t(s,k,a),t(s,cs,a),t(s,A,a),t(s,Ts,a),J(Q,s,a),t(s,os,a),J(R,s,a),t(s,Us,a),t(s,x,a),t(s,js,a),J(G,s,a),t(s,hs,a),t(s,H,a),ws=!0},p:As,i(s){ws\|\|(m(j.$$.fragment,s),m(w.$$.fragment,s),m(d.$$.fragment,s),m(C.$$.fragment,s),m(g.$$.fragment,s),m($.$$.fragment,s),m(W.$$.fragment,s),m(Z.$$.fragment,s),m(v.$$.fragment,s),m(V.$$.fragment,s),m(Q.$$.fragment,s),m(R.$$.fragment,s),m(G.$$.fragment,s),ws=!0)},o(s){c(j.$$.fragment,s),c(w.$$.fragment,s),c(d.$$.fragment,s),c(C.$$.fragment,s),c(g.$$.fragment,s),c($.$$.fragment,s),c(W.$$.fragment,s),c(Z.$$.fragment,s),c(v.$$.fragment,s),c(V.$$.fragment,s),c(Q.$$.fragment,s),c(R.$$.fragment,s),c(G.$$.fragment,s),ws=!1},d(s){s&&(l(S),l(N),l(z),l(F),l(h),l(E),l(D),l(u),l(L),l(I),l(P),l(K),l(O),l(f),l(ss),l(b),l(as),l(ls),l(_),l(ts),l(es),l(q),l(ns),l(ps),l(X),l(Ms),l(is),l(B),l(ys),l(rs),l(Y),l(Js),l(ms),l(k),l(cs),l(A),l(Ts),l(os),l(Us),l(x),l(js),l(hs),l(H)),l(o),T(j,s),T(w,s),T(d,s),T(C,s),T(g,s),T($,s),T(W,s),T(Z,s),T(v,s),T(V,s),T(Q,s),T(R,s),T(G,s)}}}const zs='{"title":"Trainer API를 사용한 하이퍼파라미터 탐색","local":"hyperparameter-search-using-trainer-api","sections":[{"title":"하이퍼파라미터 탐색 백엔드","local":"hyperparameter-search-backend","sections":[],"depth":2},{"title":"예제에서 하이퍼파라미터 탐색을 활성화하는 방법","local":"how-to-enable-hyperparameter-search-in-example","sections":[],"depth":2},{"title":"DDP 미세 조정을 위한 하이퍼파라미터 탐색","local":"hyperparameter-search-for-ddp-finetune","sections":[],"depth":2}],"depth":1}';function Fs(Is){return Qs(()=>{new URLSearchParams(window.location.search).get("fw")}),[]}class Ks extends Rs{constructor(o){super(),xs(this,o,Fs,Ss,ks,{})}}export{Ks as component};

Xet Storage Details

Size:: 21 kB
Xet hash:: 5c10a785e2b736e474c04db2b89bcc818b95f8122be7192a40604a7d9706888e

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.