Hugging Face's logo

Buckets:
hf-doc-build
/
doc

Files
xet
hf-doc-build/doc / diffusers /v0.43.0 /ko /api /pipelines /echo.html
mishig's picture
mishig
download
raw
27.5 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Echo&quot;,&quot;local&quot;:&quot;echo&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;추론&quot;,&quot;local&quot;:&quot;inference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;EchoModularPipeline&quot;,&quot;local&quot;:&quot;diffusers.EchoModularPipeline&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;EchoBlocks&quot;,&quot;local&quot;:&quot;diffusers.EchoBlocks&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/>
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/entry/start.1gaXe03k.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/BObs6oQO.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/DPnyVmXs.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/entry/app.BwdYRbCn.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/DSGUT_OW.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/C045MWWx.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/6tuJ3nyi.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/nodes/0.CornNeQR.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/IaDvMxLQ.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/nodes/177.B7YYsNYX.js" rel="modulepreload">
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/bKdwVYpU.js" rel="modulepreload">
<!--rbrdnd--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Echo&quot;,&quot;local&quot;:&quot;echo&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Inference&quot;,&quot;local&quot;:&quot;inference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;EchoModularPipeline&quot;,&quot;local&quot;:&quot;diffusers.EchoModularPipeline&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;EchoBlocks&quot;,&quot;local&quot;:&quot;diffusers.EchoBlocks&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/><!---->
<link href="/docs/diffusers/v0.43.0/en/_app/immutable/assets/0.tn0RQdqM.css" rel="modulepreload"> <!--[--><!--[0--><!--[--><!--[0--><!--[--><p></p> <!--[0--><h1 class="relative group"><a id="echo" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#echo"><span><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></span></a> <span><!---->Echo<!----></span></h1><!--]--><!----> <p><a href="https://github.com/jd-opensource/JoyAI-Echo" rel="nofollow">Echo</a>는 장시간 영상 생성 모델입니다. 선택적인 깨끗한 첫 프레임, 순서화된 이미지/오디오 메모리 슬롯, 그리고 확률적 소수 스텝 Distribution Matching Distillation (DMD) 샘플러를 지원합니다.
이 파이프라인은 동기화된 영상과 오디오를 생성합니다.</p> <p>Echo는 <a href="/docs/diffusers/v0.43.0/ko/api/modular_diffusers/pipeline#diffusers.ModularPipeline">ModularPipeline</a>으로 구현되어 있어 텍스트 인코딩, 메모리 조건화, 확률적 DMD 디노이징, 디코딩 블록을 완전한 워크플로로 실행하거나 독립적으로 조합할 수 있습니다.</p> <!--[1--><h2 class="relative group"><a id="inference" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#inference"><span><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></span></a> <span><!---->추론<!----></span></h2><!--]--><!----> <p>공식 <a href="https://huggingface.co/jdopensource/JoyAI-Echo" rel="nofollow">jdopensource/JoyAI-Echo</a> 체크포인트를 직접 로드합니다.
이 모델은 <a href="https://huggingface.co/google/gemma-3-12b-it" rel="nofollow">google/gemma-3-12b-it</a>의 텍스트 인코더와 토크나이저를 사용합니다.
Gemma 저장소는 게이트가 적용되어 있으므로, 파이프라인을 로드하기 전에 해당 라이선스에 동의하고 Hugging Face로 인증해야 합니다.</p> <p>공개된 모델은 긴 영상 예시에서 241프레임을 사용합니다. 출력 컨테이너의 프레임 레이트와 무관하게, 영상 RoPE 좌표는 학습 시 사용된 24fps의 비율을 유지합니다.</p> <div class="code-block relative "><div class="absolute top-2.5 right-4"><button class="inline-flex items-center relative text-sm focus:text-green-500 cursor-pointer focus:outline-none transition duration-200 ease-in-out opacity-0 mx-0.5 text-gray-600 " title="code excerpt" type="button"><svg xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M28,10V28H10V10H28m0-2H10a2,2,0,0,0-2,2V28a2,2,0,0,0,2,2H28a2,2,0,0,0,2-2V10a2,2,0,0,0-2-2Z" transform="translate(0)"></path><path d="M4,18H2V4A2,2,0,0,1,4,2H18V4H4Z" transform="translate(0)"></path><rect fill="none" width="32" height="32"></rect></svg><!----> <div class=" absolute pointer-events-none transition-opacity bg-black text-white py-1 px-2 leading-tight rounded font-normal shadow left-1/2 top-full transform -translate-x-1/2 translate-y-2 opacity-0 "><div class="absolute bottom-full left-1/2 transform -translate-x-1/2 w-0 h-0 border-black border-4 border-t-0" style="border-left-color: transparent; border-right-color: transparent;"></div> 복사됨</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">import</span> torchaudio
<span class="hljs-keyword">from</span> PIL <span class="hljs-keyword">import</span> Image
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> ComponentsManager, ModularPipeline
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> encode_video
manager = ComponentsManager()
pipe = ModularPipeline.from_pretrained(<span class="hljs-string">&quot;jdopensource/JoyAI-Echo&quot;</span>, components_manager=manager)
pipe.load_components(dtype={<span class="hljs-string">&quot;default&quot;</span>: torch.bfloat16, <span class="hljs-string">&quot;audio_vae&quot;</span>: torch.float32})
manager.enable_auto_cpu_offload(device=<span class="hljs-string">&quot;cuda&quot;</span>)
pipe.vae.enable_tiling()
first_frame = Image.<span class="hljs-built_in">open</span>(<span class="hljs-string">&quot;first_frame.png&quot;</span>).convert(<span class="hljs-string">&quot;RGB&quot;</span>)
memory_images = [Image.<span class="hljs-built_in">open</span>(path).convert(<span class="hljs-string">&quot;RGB&quot;</span>) <span class="hljs-keyword">for</span> path <span class="hljs-keyword">in</span> [<span class="hljs-string">&quot;memory_0.png&quot;</span>, <span class="hljs-string">&quot;memory_1.png&quot;</span>]]
memory_audio_with_rates = [torchaudio.load(path) <span class="hljs-keyword">for</span> path <span class="hljs-keyword">in</span> [<span class="hljs-string">&quot;memory_0.wav&quot;</span>, <span class="hljs-string">&quot;memory_1.wav&quot;</span>]]
memory_audio = [waveform <span class="hljs-keyword">for</span> waveform, _ <span class="hljs-keyword">in</span> memory_audio_with_rates]
memory_audio_rates = [sample_rate <span class="hljs-keyword">for</span> _, sample_rate <span class="hljs-keyword">in</span> memory_audio_with_rates]
output = pipe(
prompt=<span class="hljs-string">&quot;A cinematic dialogue scene in a quiet cafe.&quot;</span>,
image=first_frame,
memory_images=memory_images,
memory_audio_waveforms=memory_audio,
memory_audio_sample_rates=memory_audio_rates,
width=<span class="hljs-number">1280</span>,
height=<span class="hljs-number">736</span>,
num_frames=<span class="hljs-number">241</span>,
frame_rate=<span class="hljs-number">25.0</span>,
model_frame_rate=<span class="hljs-number">24.0</span>,
generator=torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">42</span>),
output_type=<span class="hljs-string">&quot;np&quot;</span>,
output=[<span class="hljs-string">&quot;videos&quot;</span>, <span class="hljs-string">&quot;audio&quot;</span>],
)
encode_video(
output[<span class="hljs-string">&quot;videos&quot;</span>][<span class="hljs-number">0</span>],
fps=<span class="hljs-number">25</span>,
audio=output[<span class="hljs-string">&quot;audio&quot;</span>][<span class="hljs-number">0</span>].<span class="hljs-built_in">float</span>().cpu(),
audio_sample_rate=pipe.vocoder.config.output_sampling_rate,
output_path=<span class="hljs-string">&quot;echo.mp4&quot;</span>,
)<!----></pre></div><!----> <p>기본 DMD sigma 스케줄은 공개된 8단계 스케줄입니다. 모든 단계에서 <code>x0</code>를 예측하고 다음 sigma에서 새로운 가우시안 노이즈로 재노이즈링을 수행하므로, 시드가 설정된 <code>torch.Generator</code>가 초기 노이즈와 모든 중간 재노이즈링을 제어합니다.</p> <p>Raw 오디오-메모리 인코딩에는 <code>torchaudio</code>가 필요합니다. 참조와 동일한 조건을 유지하려면, 위와 같이 <code>audio_vae</code>를 FP32로 유지하세요.
모듈형 워크플로에서는 VAE 인코더의 정규화되고 언팩된 텐서를 캐시할 수 있습니다. 비디오 라텐트는 <code>(batch, channels, frames, height, width)</code> 형태를, 오디오 라텐트는 <code>(batch, channels, time, mel_bins)</code> 형태를 가집니다.
핵심 <code>denoise</code> 블록은 해당 텐서를 팩하고, <code>num_videos_per_prompt</code>에 맞춰 조건을 확장한 후, 디노이즈된 출력을 동일한 VAE 형태로 다시 언팩합니다. 초기 <code>latents</code>와 <code>audio_latents</code>도 동일한 언팩된 형태로 전달하세요.
디코더는 정규화된 VAE 텐서를 받아, 디코딩을 수행하기 직전에 즉시 비정규화합니다. <code>output=["latents", "audio_latents"]</code>를 사용하면 정규화된 VAE 텐서가 반환됩니다. <code>output_type="latent"</code>와 <code>output=["videos", "audio"]</code>를 함께 사용하면 디코딩 없이 비정규화된 VAE 텐서를 얻을 수 있습니다.</p> <!--[1--><h2 class="relative group"><a id="diffusers.EchoModularPipeline" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#diffusers.EchoModularPipeline"><span><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></span></a> <span><!---->EchoModularPipeline<!----></span></h2><!--]--><!----> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.EchoModularPipeline"><!----><h3 class="!m-0"><span class="flex-1 break-all md:text-lg bg-gradient-to-r px-2.5 py-1.5 rounded-xl from-indigo-50/70 to-white dark:from-gray-900 dark:to-gray-950 dark:text-indigo-300 text-indigo-700"><svg class="mr-1.5 text-indigo-500 inline-block -mt-0.5" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width=".8em" height=".8em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 24 24"><path class="uim-quaternary" d="M20.23 7.24L12 12L3.77 7.24a1.98 1.98 0 0 1 .7-.71L11 2.76c.62-.35 1.38-.35 2 0l6.53 3.77c.29.173.531.418.7.71z" opacity=".25" fill="currentColor"></path><path class="uim-tertiary" d="M12 12v9.5a2.09 2.09 0 0 1-.91-.21L4.5 17.48a2.003 2.003 0 0 1-1-1.73v-7.5a2.06 2.06 0 0 1 .27-1.01L12 12z" opacity=".5" fill="currentColor"></path><path class="uim-primary" d="M20.5 8.25v7.5a2.003 2.003 0 0 1-1 1.73l-6.62 3.82c-.275.13-.576.198-.88.2V12l8.23-4.76c.175.308.268.656.27 1.01z" fill="currentColor"></path></svg><span class="font-light">클래스</span> <span class="font-medium">diffusers.</span><span class="font-semibold">EchoModularPipeline</span></span></h3><!----> <a id="diffusers.EchoModularPipeline" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.EchoModularPipeline"><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/modular_pipelines/echo/modular_pipeline.py#L27" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">blocks<span class="opacity-60">: diffusers.modular_pipelines.modular_pipeline.ModularPipelineBlocks | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">pretrained_model_name_or_path<span class="opacity-60">: str | os.PathLike | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">components_manager<span class="opacity-60">: diffusers.modular_pipelines.components_manager.ComponentsManager | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">collection<span class="opacity-60">: str | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">workflow<span class="opacity-60">: str | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">modular_config_dict<span class="opacity-60">: dict[str, typing.Any] | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">config_dict<span class="opacity-60">: dict[str, typing.Any] | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">**kwargs<span class="opacity-60"></span></span></span><!--]--> <span>)</span> <!--[-1--><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>Echo 장시간 비디오 DMD 체크포인트용 ModularPipeline.</p></div> <!--[1--><h2 class="relative group"><a id="diffusers.EchoBlocks" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#diffusers.EchoBlocks"><span><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></span></a> <span><!---->EchoBlocks<!----></span></h2><!--]--><!----> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.EchoBlocks"><!----><h3 class="!m-0"><span class="flex-1 break-all md:text-lg bg-gradient-to-r px-2.5 py-1.5 rounded-xl from-indigo-50/70 to-white dark:from-gray-900 dark:to-gray-950 dark:text-indigo-300 text-indigo-700"><svg class="mr-1.5 text-indigo-500 inline-block -mt-0.5" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width=".8em" height=".8em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 24 24"><path class="uim-quaternary" d="M20.23 7.24L12 12L3.77 7.24a1.98 1.98 0 0 1 .7-.71L11 2.76c.62-.35 1.38-.35 2 0l6.53 3.77c.29.173.531.418.7.71z" opacity=".25" fill="currentColor"></path><path class="uim-tertiary" d="M12 12v9.5a2.09 2.09 0 0 1-.91-.21L4.5 17.48a2.003 2.003 0 0 1-1-1.73v-7.5a2.06 2.06 0 0 1 .27-1.01L12 12z" opacity=".5" fill="currentColor"></path><path class="uim-primary" d="M20.5 8.25v7.5a2.003 2.003 0 0 1-1 1.73l-6.62 3.82c-.275.13-.576.198-.88.2V12l8.23-4.76c.175.308.268.656.27 1.01z" fill="currentColor"></path></svg><span class="font-light">클래스</span> <span class="font-medium">diffusers.</span><span class="font-semibold">EchoBlocks</span></span></h3><!----> <a id="diffusers.EchoBlocks" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.EchoBlocks"><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/v0.43.0/src/diffusers/modular_pipelines/echo/modular_blocks_echo.py#L220" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><!--]--> <span>)</span> <!--[-1--><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>Echo는 깔끔한 첫 프레임 조건부 설정, 정렬된 이미지/오디오 메모리 슬롯, 그리고 확률적 DMD 디노이징을 활용한 참조-영상 생성을 수행합니다.</p> <p>구성 요소:
text_encoder (<code>PreTrainedModel</code>) tokenizer (<code>PreTrainedTokenizerBase</code>) connectors (<code>LTX2TextConnectors</code>) vae
(<code>AutoencoderKLLTX2Video</code>) audio_vae (<code>AutoencoderKLLTX2Audio</code>) transformer (<code>LTX2VideoTransformer3DModel</code>)
video_processor (<code>VideoProcessor</code>) vocoder (<code>LTX2Vocoder</code>)</p> <p>입력:
prompt (<code>str</code>):
이미지 생성을 안내하는 프롬프트 또는 프롬프트 목록입니다.
max_sequence_length (<code>int</code>, <em>선택 사항</em>, 기본값 1024):
프롬프트 인코딩을 위한 최대 시퀀스 길이입니다.
image (<code>Image | Tensor</code>, <em>선택 사항</em>):
깨끗한 참조 조건으로 사용되는 선택적 단일 첫 번째 프레임입니다.
memory_images (<code>list</code>, <em>선택 사항</em>):
Echo 메모리 슬롯당 하나씩 정렬된 참조 이미지 목록입니다.
memory_audio_waveforms (<code>list</code>, <em>선택 사항</em>):
<code>(channels, samples)</code> 텐서 형식으로 정렬된 메모리 웨이브폼입니다. 9.62초보다 긴 입력은 가장 높은 응답을 보이는 윈도우로 잘립니다. 무음 슬롯의 경우 <code>None</code>을 사용합니다.
memory_audio_sample_rates (<code>int | list</code>, <em>선택 사항</em>):
모든 메모리 웨이브폼이 공유하는 샘플링 레이트 또는 슬롯별 하나의 레이트입니다.
height (<code>int</code>, <em>선택 사항</em>, 기본값 512):
생성된 이미지의 픽셀 단위 높이입니다.
width (<code>int</code>, <em>선택 사항</em>, 기본값 704):
생성된 이미지의 픽셀 단위 너비입니다.
model_frame_rate (<code>float</code>, <em>선택 사항</em>, 기본값 24.0):
Echo 비디오 RoPE 좌표에 사용되는 학습 시 프레임 레이트입니다.
memory_position_offset (<code>float</code>, <em>선택 사항</em>, 기본값 500.0):
첫 번째 메모리 슬롯에 할당되는 시간 중심값입니다.
memory_position_slot_stride (<code>float</code>, <em>선택 사항</em>, 기본값 50.0):
연속된 메모리 슬롯 중심 사이의 시간 간격입니다.
num_videos_per_prompt (<code>int</code>, <em>선택 사항</em>, 기본값 1):
프롬프트당 생성할 비디오 수입니다.
num_frames (<code>int</code>, <em>선택 사항</em>, 기본값 241):
생성할 픽셀 프레임 수입니다. <code>1 + k * vae_temporal_compression_ratio</code> 형태여야 합니다.
frame_rate (<code>float</code>, <em>선택 사항</em>, 기본값 25.0):
생성된 비디오의 프레임 레이트입니다.
latents (<code>Tensor</code>, <em>선택 사항</em>):
VAE 형식 (B, C, F, H, W)의 선택적 초기 비디오 노이즈입니다.
audio_latents (<code>Tensor</code>, <em>선택 사항</em>):
VAE 형식 (B, C, L, M)의 선택적 초기 오디오 노이즈입니다.
generator (<code>Generator</code>, <em>선택 사항</em>):
결정론적 생성을 위한 Torch 생성기입니다.
sigmas (<code>list | tuple</code>):
종단부 0을 포함한 DMD 시그마 스케줄입니다.
attention_kwargs (<code>dict</code>, <em>선택 사항</em>):
어텐션 프로세서에 전달할 추가 kwargs입니다.
output_type (<code>str</code>, <em>선택 사항</em>, 기본값 pil):
출력 형식: ‘pil’, ‘np’, ‘pt’.
decode_timestep (<code>None</code>, <em>선택 사항</em>, 기본값 0.0):
최종 latents를 디코딩하는 데 사용되는 타임스텝입니다.
decode_noise_scale (<code>None</code>, <em>선택 사항</em>):
디코딩 타임스텝에서 적용되는 노이즈 보간 계수입니다.</p> <p>Outputs:
videos (<code>list</code>):
생성된 비디오.
audio (<code>Tensor</code>):
생성된 오디오 웨이브폼.</p></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/diffusers/blob/main/docs/source/en/api/pipelines/echo.md" target="_blank"><svg class="mr-1" xmlns="http://www.w3.org/2000/svg" aria-hidden="true" fill="currentColor" focusable="false" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 32 32"><path d="M31,16l-7,7l-1.41-1.41L28.17,16l-5.58-5.59L24,9l7,7z"></path><path d="M1,16l7-7l1.41,1.41L3.83,16l5.58,5.59L8,23l-7-7z"></path><path d="M12.419,25.484L17.639,6.552l1.932,0.518L14.351,26.002z"></path></svg><!----> <span><span class="underline">Update</span> on GitHub</span></a><!----> <p></p><!--]--><!----><!--]--><!--]--><!--]--> <!--[-1--><!--]--><!--]-->
<script>
{
__sveltekit_ywq9g6 = {
base: "/docs/diffusers/v0.43.0/en",
assets: "/docs/diffusers/v0.43.0/en"
};
const element = document.currentScript.parentElement;
Promise.all([
import("/docs/diffusers/v0.43.0/en/_app/immutable/entry/start.1gaXe03k.js"),
import("/docs/diffusers/v0.43.0/en/_app/immutable/entry/app.BwdYRbCn.js")
]).then(([kit, app]) => {
kit.start(app, element, {
node_ids: [0, 177],
data: [null,null],
form: null,
error: null
});
});
}
</script>

Xet Storage Details

Size:
27.5 kB
·
Xet hash:
2a0e8eb746e8e5069b44baeb9bf807fe0fe862be8e4578567606929baf777719

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