diffusers/pr_14169/ko/_app/immutable/nodes/8.D1CMCj9E.js

import"../chunks/DsnmJJEf.js";import{i as B,h as C,C as _,H as e,a as s,E as R,s as k}from"../chunks/okyp_udp.js";import{p as E,o as v,s as l,f as A,a as W,b as Y,c as N,n as x}from"../chunks/CSNmR2bO.js";const Q='{"title":"스케줄러","local":"스케줄러","sections":[{"title":"파이프라인 불러오기","local":"파이프라인-불러오기","sections":[],"depth":2},{"title":"스케줄러 액세스","local":"스케줄러-액세스","sections":[],"depth":2},{"title":"스케줄러 교체하기","local":"스케줄러-교체하기","sections":[],"depth":2},{"title":"스케줄러들 비교해보기","local":"스케줄러들-비교해보기","sections":[],"depth":2}],"depth":1}';var D=N('<meta name="hf:doc:metadata"/>'),z=N('<p></p> <!> <!> <p>diffusion 파이프라인은 diffusion 모델, 스케줄러 등의 컴포넌트들로 구성됩니다. 그리고 파이프라인 안의 일부 컴포넌트를 다른 컴포넌트로 교체하는 식의 커스터마이징 역시 가능합니다.  이와 같은 컴포넌트 커스터마이징의 가장 대표적인 예시가 바로 <a href="../api/schedulers/overview">스케줄러</a>를 교체하는 것입니다.</p> <p>스케쥴러는 다음과 같이 diffusion 시스템의 전반적인 디노이징 프로세스를 정의합니다.</p> <ul><li>디노이징 스텝을 얼마나 가져가야 할까?</li> <li>확률적으로(stochastic) 혹은 확정적으로(deterministic)?</li> <li>디노이징 된 샘플을 찾아내기 위해 어떤 알고리즘을 사용해야 할까?</li></ul> <p>이러한 프로세스는 다소 난해하고, 디노이징 속도와 디노이징 퀄리티 사이의 트레이드 오프를 정의해야 하는 문제가 될 수 있습니다. 주어진 파이프라인에 어떤 스케줄러가 가장 적합한지를 정량적으로 판단하는 것은 매우 어려운 일입니다. 이로 인해 일단 해당 스케줄러를 직접 사용하여, 생성되는 이미지를 직접 눈으로 보며, 정성적으로 성능을 판단해보는 것이 추천되곤 합니다.</p> <!> <p>먼저 스테이블 diffusion 파이프라인을 불러오도록 해보겠습니다. 물론 스테이블 diffusion을 사용하기 위해서는, 허깅페이스 허브에 등록된 사용자여야 하며, 관련 <a href="https://huggingface.co/stable-diffusion-v1-5/stable-diffusion-v1-5" rel="nofollow">라이센스</a>에 동의해야 한다는 점을 잊지 말아주세요.</p> <p><em>역자 주: 다만, 현재 신규로 생성한 허깅페이스 계정에 대해서는 라이센스 동의를 요구하지 않는 것으로 보입니다!</em></p> <!> <p>다음으로, GPU로 이동합니다.</p> <!> <!> <p>스케줄러는 언제나 파이프라인의 컴포넌트로서 존재하며, 일반적으로 파이프라인 인스턴스 내에 <code>scheduler</code>라는 이름의 속성(property)으로 정의되어 있습니다.</p> <!> <p><strong>Output</strong>:</p> <!> <p>출력 결과를 통해, 우리는 해당 스케줄러가 <code>PNDMScheduler</code>의 인스턴스라는 것을 알 수 있습니다. 이제 <code>PNDMScheduler</code>와 다른 스케줄러들의 성능을 비교해보도록 하겠습니다. 먼저 테스트에 사용할 프롬프트를 다음과 같이 정의해보도록 하겠습니다.</p> <!> <p>다음으로 유사한 이미지 생성을 보장하기 위해서, 다음과 같이 랜덤시드를 고정해주도록 하겠습니다.</p> <!> <p align="center"><br/> <img src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_pndm.png" width="400"/> <br/></p> <!> <p>다음으로 파이프라인의 스케줄러를 다른 스케줄러로 교체하는 방법에 대해 알아보겠습니다. 모든 스케줄러는 <code>SchedulerMixin.compatibles</code>라는 속성(property)을 갖고 있습니다. 해당 속성은 <strong>호환 가능한</strong> 스케줄러들에 대한 정보를 담고 있습니다.</p> <!> <p><strong>Output</strong>:</p> <!> <p>호환되는 스케줄러들을 살펴보면 아래와 같습니다.</p> <ul><li><code>LMSDiscreteScheduler</code>,</li> <li><code>DDIMScheduler</code>,</li> <li><code>DPMSolverMultistepScheduler</code>,</li> <li><code>EulerDiscreteScheduler</code>,</li> <li><code>PNDMScheduler</code>,</li> <li><code>DDPMScheduler</code>,</li> <li><code>EulerAncestralDiscreteScheduler</code>.</li></ul> <p>앞서 정의했던 프롬프트를 사용해서 각각의 스케줄러들을 비교해보도록 하겠습니다.</p> <p>먼저 파이프라인 안의 스케줄러를 바꾸기 위해 <code>ConfigMixin.config</code> 속성과 <code>ConfigMixin.from_config()</code> 메서드를 활용해보려고 합니다.</p> <!> <p><strong>Output</strong>:</p> <!> <p>기존 스케줄러의 config를 호환 가능한 다른 스케줄러에 이식하는 것 역시 가능합니다.</p> <p>다음 예시는 기존 스케줄러(<code>pipeline.scheduler</code>)를 다른 종류의 스케줄러(<code>DDIMScheduler</code>)로 바꾸는 코드입니다. 기존 스케줄러가 갖고 있던 config를 <code>.from_config</code> 메서드의 인자로 전달하는 것을 확인할 수 있습니다.</p> <!> <p>이제 파이프라인을 실행해서 두 스케줄러 사이의 생성된 이미지의 퀄리티를 비교해봅시다.</p> <!> <p align="center"><br/> <img src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_ddim.png" width="400"/> <br/></p> <!> <p>지금까지는 <code>PNDMScheduler</code>와 <code>DDIMScheduler</code> 스케줄러를 실행해보았습니다. 아직 비교해볼 스케줄러들이 더 많이 남아있으니 계속 비교해보도록 하겠습니다.</p> <p><code>LMSDiscreteScheduler</code>을 일반적으로 더 좋은 결과를 보여줍니다.</p> <!> <p align="center"><br/> <img src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_lms.png" width="400"/> <br/></p> <p><code>EulerDiscreteScheduler</code>와 <code>EulerAncestralDiscreteScheduler</code> 고작 30번의 inference step만으로도 높은 퀄리티의 이미지를 생성하는 것을 알 수 있습니다.</p> <!> <p align="center"><br/> <img src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_euler_discrete.png" width="400"/> <br/></p> <!> <p align="center"><br/> <img src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_euler_ancestral.png" width="400"/> <br/></p> <p>지금 이 문서를 작성하는 현시점 기준에선, <code>DPMSolverMultistepScheduler</code>가 시간 대비 가장 좋은 품질의 이미지를 생성하는 것 같습니다. 20번 정도의 스텝만으로도 실행될 수 있습니다.</p> <!> <p align="center"><br/> <img src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_dpm.png" width="400"/> <br/></p> <p>보시다시피 생성된 이미지들은 매우 비슷하고, 비슷한 퀄리티를 보이는 것 같습니다. 실제로 어떤 스케줄러를 선택할 것인가는 종종 특정 이용 사례에 기반해서 결정되곤 합니다. 결국 여러 종류의 스케줄러를 직접 실행시켜보고 눈으로 직접 비교해서 판단하는 게 좋은 선택일 것 같습니다.</p> <!> <p></p>',1);function L(I,X){E(X,!1),v(()=>{new URLSearchParams(window.location.search).get("fw")}),B();var a=z();C("alcvz4",T=>{var V=D();k(V,"content",Q),W(T,V)});var n=l(A(a),2);_(n,{containerStyle:"float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"});var c=l(n,2);e(c,{title:"스케줄러",local:"스케줄러",headingTag:"h1"});var p=l(c,10);e(p,{title:"파이프라인 불러오기",local:"파이프라인-불러오기",headingTag:"h2"});var r=l(p,6);s(r,{code:"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",highlighted:`<span class="hljs-keyword">from</span> huggingface_hub <span class="hljs-keyword">import</span> login
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline
<span class="hljs-keyword">import</span> torch

<span class="hljs-comment"># first we need to login with our access token</span>
login()

<span class="hljs-comment"># Now we can download the pipeline</span>
pipeline = DiffusionPipeline.from_pretrained(<span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>, torch_dtype=torch.float16)`,lang:"python",wrap:!1});var i=l(r,4);s(i,{code:"cGlwZWxpbmUudG8oJTIyY3VkYSUyMik=",highlighted:'pipeline.to(<span class="hljs-string">&quot;cuda&quot;</span>)',lang:"python",wrap:!1});var o=l(i,2);e(o,{title:"스케줄러 액세스",local:"스케줄러-액세스",headingTag:"h2"});var t=l(o,4);s(t,{code:"cGlwZWxpbmUuc2NoZWR1bGVy",highlighted:"pipeline.scheduler",lang:"python",wrap:!1});var d=l(t,4);s(d,{code:"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",highlighted:`PNDMScheduler {
  <span class="hljs-string">&quot;_class_name&quot;</span>: <span class="hljs-string">&quot;PNDMScheduler&quot;</span>,
  <span class="hljs-string">&quot;_diffusers_version&quot;</span>: <span class="hljs-string">&quot;0.8.0.dev0&quot;</span>,
  <span class="hljs-string">&quot;beta_end&quot;</span>: <span class="hljs-number">0.012</span>,
  <span class="hljs-string">&quot;beta_schedule&quot;</span>: <span class="hljs-string">&quot;scaled_linear&quot;</span>,
  <span class="hljs-string">&quot;beta_start&quot;</span>: <span class="hljs-number">0.00085</span>,
  <span class="hljs-string">&quot;clip_sample&quot;</span>: <span class="hljs-literal">false</span>,
  <span class="hljs-string">&quot;num_train_timesteps&quot;</span>: <span class="hljs-number">1000</span>,
  <span class="hljs-string">&quot;set_alpha_to_one&quot;</span>: <span class="hljs-literal">false</span>,
  <span class="hljs-string">&quot;skip_prk_steps&quot;</span>: <span class="hljs-literal">true</span>,
  <span class="hljs-string">&quot;steps_offset&quot;</span>: <span class="hljs-number">1</span>,
  <span class="hljs-string">&quot;trained_betas&quot;</span>: <span class="hljs-literal">null</span>
}`,lang:"",wrap:!1});var M=l(d,4);s(M,{code:"cHJvbXB0JTIwJTNEJTIwJTIyQSUyMHBob3RvZ3JhcGglMjBvZiUyMGFuJTIwYXN0cm9uYXV0JTIwcmlkaW5nJTIwYSUyMGhvcnNlJTIwb24lMjBNYXJzJTJDJTIwaGlnaCUyMHJlc29sdXRpb24lMkMlMjBoaWdoJTIwZGVmaW5pdGlvbi4lMjI=",highlighted:'prompt = <span class="hljs-string">&quot;A photograph of an astronaut riding a horse on Mars, high resolution, high definition.&quot;</span>',lang:"python",wrap:!1});var h=l(M,4);s(h,{code:"Z2VuZXJhdG9yJTIwJTNEJTIwdG9yY2guR2VuZXJhdG9yKGRldmljZSUzRCUyMmN1ZGElMjIpLm1hbnVhbF9zZWVkKDgpJTBBaW1hZ2UlMjAlM0QlMjBwaXBlbGluZShwcm9tcHQlMkMlMjBnZW5lcmF0b3IlM0RnZW5lcmF0b3IpLmltYWdlcyU1QjAlNUQlMEFpbWFnZQ==",highlighted:`generator = torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">8</span>)
image = pipeline(prompt, generator=generator).images[<span class="hljs-number">0</span>]
image`,lang:"python",wrap:!1});var u=l(h,4);e(u,{title:"스케줄러 교체하기",local:"스케줄러-교체하기",headingTag:"h2"});var y=l(u,4);s(y,{code:"cGlwZWxpbmUuc2NoZWR1bGVyLmNvbXBhdGlibGVz",highlighted:"pipeline.scheduler.compatibles",lang:"python",wrap:!1});var m=l(y,4);s(m,{code:"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",highlighted:`[<span class="hljs-keyword">diffusers.schedulers.scheduling_lms_discrete.LMSDiscreteScheduler,
</span> <span class="hljs-keyword">diffusers.schedulers.scheduling_ddim.DDIMScheduler,
</span> <span class="hljs-keyword">diffusers.schedulers.scheduling_dpmsolver_multistep.DPMSolverMultistepScheduler,
</span> <span class="hljs-keyword">diffusers.schedulers.scheduling_euler_discrete.EulerDiscreteScheduler,
</span> <span class="hljs-keyword">diffusers.schedulers.scheduling_pndm.PNDMScheduler,
</span> <span class="hljs-keyword">diffusers.schedulers.scheduling_ddpm.DDPMScheduler,
</span> <span class="hljs-keyword">diffusers.schedulers.scheduling_euler_ancestral_discrete.EulerAncestralDiscreteScheduler]</span>`,lang:"",wrap:!1});var Z=l(m,10);s(Z,{code:"cGlwZWxpbmUuc2NoZWR1bGVyLmNvbmZpZw==",highlighted:"pipeline.scheduler.config",lang:"python",wrap:!1});var g=l(Z,4);s(g,{code:"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",highlighted:`FrozenDict([(&#x27;num_train_timesteps&#x27;, <span class="hljs-number">1000</span>),
            (&#x27;beta_start&#x27;, <span class="hljs-number">0.00085</span>),
            (&#x27;beta_end&#x27;, <span class="hljs-number">0.012</span>),
            (&#x27;beta_schedule&#x27;, &#x27;scaled_linear&#x27;),
            (&#x27;trained_betas&#x27;, None),
            (&#x27;skip_prk_steps&#x27;, True),
            (&#x27;set_alpha_to_one&#x27;, False),
            (&#x27;steps_offset&#x27;, <span class="hljs-number">1</span>),
            (&#x27;_class_name&#x27;, &#x27;PNDMScheduler&#x27;),
            (&#x27;_diffusers_version&#x27;, &#x27;<span class="hljs-number">0.8</span>.<span class="hljs-number">0</span>.dev0&#x27;),
            (&#x27;clip_sample&#x27;, False)])`,lang:"",wrap:!1});var j=l(g,6);s(j,{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMERESU1TY2hlZHVsZXIlMEElMEFwaXBlbGluZS5zY2hlZHVsZXIlMjAlM0QlMjBERElNU2NoZWR1bGVyLmZyb21fY29uZmlnKHBpcGVsaW5lLnNjaGVkdWxlci5jb25maWcp",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DDIMScheduler

pipeline.scheduler = DDIMScheduler.from_config(pipeline.scheduler.config)`,lang:"python",wrap:!1});var U=l(j,4);s(U,{code:"Z2VuZXJhdG9yJTIwJTNEJTIwdG9yY2guR2VuZXJhdG9yKGRldmljZSUzRCUyMmN1ZGElMjIpLm1hbnVhbF9zZWVkKDgpJTBBaW1hZ2UlMjAlM0QlMjBwaXBlbGluZShwcm9tcHQlMkMlMjBnZW5lcmF0b3IlM0RnZW5lcmF0b3IpLmltYWdlcyU1QjAlNUQlMEFpbWFnZQ==",highlighted:`generator = torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">8</span>)
image = pipeline(prompt, generator=generator).images[<span class="hljs-number">0</span>]
image`,lang:"python",wrap:!1});var J=l(U,4);e(J,{title:"스케줄러들 비교해보기",local:"스케줄러들-비교해보기",headingTag:"h2"});var b=l(J,6);s(b,{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMExNU0Rpc2NyZXRlU2NoZWR1bGVyJTBBJTBBcGlwZWxpbmUuc2NoZWR1bGVyJTIwJTNEJTIwTE1TRGlzY3JldGVTY2hlZHVsZXIuZnJvbV9jb25maWcocGlwZWxpbmUuc2NoZWR1bGVyLmNvbmZpZyklMEElMEFnZW5lcmF0b3IlMjAlM0QlMjB0b3JjaC5HZW5lcmF0b3IoZGV2aWNlJTNEJTIyY3VkYSUyMikubWFudWFsX3NlZWQoOCklMEFpbWFnZSUyMCUzRCUyMHBpcGVsaW5lKHByb21wdCUyQyUyMGdlbmVyYXRvciUzRGdlbmVyYXRvcikuaW1hZ2VzJTVCMCU1RCUwQWltYWdl",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> LMSDiscreteScheduler

pipeline.scheduler = LMSDiscreteScheduler.from_config(pipeline.scheduler.config)

generator = torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">8</span>)
image = pipeline(prompt, generator=generator).images[<span class="hljs-number">0</span>]
image`,lang:"python",wrap:!1});var w=l(b,6);s(w,{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMEV1bGVyRGlzY3JldGVTY2hlZHVsZXIlMEElMEFwaXBlbGluZS5zY2hlZHVsZXIlMjAlM0QlMjBFdWxlckRpc2NyZXRlU2NoZWR1bGVyLmZyb21fY29uZmlnKHBpcGVsaW5lLnNjaGVkdWxlci5jb25maWcpJTBBJTBBZ2VuZXJhdG9yJTIwJTNEJTIwdG9yY2guR2VuZXJhdG9yKGRldmljZSUzRCUyMmN1ZGElMjIpLm1hbnVhbF9zZWVkKDgpJTBBaW1hZ2UlMjAlM0QlMjBwaXBlbGluZShwcm9tcHQlMkMlMjBnZW5lcmF0b3IlM0RnZW5lcmF0b3IlMkMlMjBudW1faW5mZXJlbmNlX3N0ZXBzJTNEMzApLmltYWdlcyU1QjAlNUQlMEFpbWFnZQ==",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> EulerDiscreteScheduler

pipeline.scheduler = EulerDiscreteScheduler.from_config(pipeline.scheduler.config)

generator = torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">8</span>)
image = pipeline(prompt, generator=generator, num_inference_steps=<span class="hljs-number">30</span>).images[<span class="hljs-number">0</span>]
image`,lang:"python",wrap:!1});var f=l(w,4);s(f,{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMEV1bGVyQW5jZXN0cmFsRGlzY3JldGVTY2hlZHVsZXIlMEElMEFwaXBlbGluZS5zY2hlZHVsZXIlMjAlM0QlMjBFdWxlckFuY2VzdHJhbERpc2NyZXRlU2NoZWR1bGVyLmZyb21fY29uZmlnKHBpcGVsaW5lLnNjaGVkdWxlci5jb25maWcpJTBBJTBBZ2VuZXJhdG9yJTIwJTNEJTIwdG9yY2guR2VuZXJhdG9yKGRldmljZSUzRCUyMmN1ZGElMjIpLm1hbnVhbF9zZWVkKDgpJTBBaW1hZ2UlMjAlM0QlMjBwaXBlbGluZShwcm9tcHQlMkMlMjBnZW5lcmF0b3IlM0RnZW5lcmF0b3IlMkMlMjBudW1faW5mZXJlbmNlX3N0ZXBzJTNEMzApLmltYWdlcyU1QjAlNUQlMEFpbWFnZQ==",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> EulerAncestralDiscreteScheduler

pipeline.scheduler = EulerAncestralDiscreteScheduler.from_config(pipeline.scheduler.config)

generator = torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">8</span>)
image = pipeline(prompt, generator=generator, num_inference_steps=<span class="hljs-number">30</span>).images[<span class="hljs-number">0</span>]
image`,lang:"python",wrap:!1});var G=l(f,6);s(G,{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMERQTVNvbHZlck11bHRpc3RlcFNjaGVkdWxlciUwQSUwQXBpcGVsaW5lLnNjaGVkdWxlciUyMCUzRCUyMERQTVNvbHZlck11bHRpc3RlcFNjaGVkdWxlci5mcm9tX2NvbmZpZyhwaXBlbGluZS5zY2hlZHVsZXIuY29uZmlnKSUwQSUwQWdlbmVyYXRvciUyMCUzRCUyMHRvcmNoLkdlbmVyYXRvcihkZXZpY2UlM0QlMjJjdWRhJTIyKS5tYW51YWxfc2VlZCg4KSUwQWltYWdlJTIwJTNEJTIwcGlwZWxpbmUocHJvbXB0JTJDJTIwZ2VuZXJhdG9yJTNEZ2VuZXJhdG9yJTJDJTIwbnVtX2luZmVyZW5jZV9zdGVwcyUzRDIwKS5pbWFnZXMlNUIwJTVEJTBBaW1hZ2U=",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DPMSolverMultistepScheduler

pipeline.scheduler = DPMSolverMultistepScheduler.from_config(pipeline.scheduler.config)

generator = torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">8</span>)
image = pipeline(prompt, generator=generator, num_inference_steps=<span class="hljs-number">20</span>).images[<span class="hljs-number">0</span>]
image`,lang:"python",wrap:!1});var S=l(G,6);R(S,{source:"https://github.com/huggingface/diffusers/blob/main/docs/source/ko/using-diffusers/schedulers.md"}),x(2),W(I,a),Y()}export{L as component};