Buckets:

hf-doc-build
/

doc

Files
xet
hf-doc-build/doc / diffusers /main /vi /using-diffusers /batched_inference.html
mishig
3 days ago
download
raw
34.7 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Suy luận theo lô&quot;,&quot;local&quot;:&quot;batch-inference&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Gom nhiều prompt thành một lô&quot;,&quot;local&quot;:&quot;batch-multiple-prompts&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Tạo nhiều hình ảnh cho mỗi prompt&quot;,&quot;local&quot;:&quot;generate-several-images-per-prompt&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Tạo một vài hình ảnh cho một vài prompt&quot;,&quot;local&quot;:&quot;generate-several-images-for-several-prompts&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Tạo các lô dữ liệu có thể tái lập&quot;,&quot;local&quot;:&quot;generate-reproducible-batches&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Xử lý các yêu cầu đồng thời&quot;,&quot;local&quot;:&quot;serve-concurrent-requests&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Các bước tiếp theo&quot;,&quot;local&quot;:&quot;next-steps&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/>
		<link href="/docs/diffusers/main/en/_app/immutable/entry/start.0S0_iEoG.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/I9eooN77.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/Bupkjr9f.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/entry/app.CPdOxCbw.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/CyvZ3FS5.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/l0QnnOED.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DnVIJfC5.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/nodes/0.DteTvmQZ.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/CoSrdQax.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/nodes/386.CyeNlhD_.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DSPDl1gz.js" rel="modulepreload">
		<!--1sdmsnt--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Batch inference&quot;,&quot;local&quot;:&quot;batch-inference&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Batch multiple prompts&quot;,&quot;local&quot;:&quot;batch-multiple-prompts&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Generate several images per prompt&quot;,&quot;local&quot;:&quot;generate-several-images-per-prompt&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Generate several images for several prompts&quot;,&quot;local&quot;:&quot;generate-several-images-for-several-prompts&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Generate reproducible batches&quot;,&quot;local&quot;:&quot;generate-reproducible-batches&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Serve concurrent requests&quot;,&quot;local&quot;:&quot;serve-concurrent-requests&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Next steps&quot;,&quot;local&quot;:&quot;next-steps&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/><!---->
		<link href="/docs/diffusers/main/en/_app/immutable/assets/0.tn0RQdqM.css" rel="modulepreload"> <!--[--><!--[0--><!--[--><!--[0--><!--[--><p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: inline-end; margin-inline-start: 10px; display: inline-flex; position: relative; z-index: 10;"><div dir="ltr" class="inline-flex rounded-md max-sm:rounded-sm"><button class="inline-flex items-center gap-1 h-7 max-sm:h-7 px-2 max-sm:px-1.5 text-sm font-medium text-gray-800 border border-r-0 rounded-l-md max-sm:rounded-l-sm border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-live="polite"><span class="inline-flex items-center justify-center rounded-md p-0.5 max-sm:p-0 hover:text-gray-800 dark:hover:text-gray-200"><svg class="sm:size-3.5 size-3" 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><!----></span> <span dir="auto">Copy page</span></button> <button class="inline-flex items-center justify-center w-6 max-sm:w-5 h-7 max-sm:h-7 disabled:pointer-events-none text-sm text-gray-500 hover:text-gray-700 dark:hover:text-white rounded-r-md max-sm:rounded-r-sm border border-l transition border-gray-200 bg-white hover:shadow-inner dark:border-gray-850 dark:bg-gray-950 dark:text-gray-200 dark:hover:bg-gray-800" aria-haspopup="menu" aria-expanded="false" aria-label="Open copy menu"><svg class="transition-transform text-gray-400 overflow-visible sm:size-3.5 size-3 rotate-0" width="1em" height="1em" viewBox="0 0 12 7" fill="none" xmlns="http://www.w3.org/2000/svg"><path d="M1 1L6 6L11 1" stroke="currentColor"></path></svg><!----></button></div> <!--[-1--><!--]--></div><!----> <!--[0--><h1 class="relative group"><a id="batch-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="#batch-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><!---->Suy luận theo lô<!----></span></h1><!--]--><!----> <p>Suy luận theo lô (batch inference) xử lý nhiều prompt cùng lúc để tăng thông lượng. Một lô dữ liệu giúp GPU hoạt động liên tục, trong khi một prompt đơn lẻ khiến GPU bị khai thác chưa hết công suất.</p> <p>Nhược điểm là độ trễ cao hơn, vì bạn phải chờ toàn bộ lô (batch) hoàn thành. Các lô lớn cũng cần nhiều bộ nhớ GPU hơn.</p> <!--[1--><h2 class="relative group"><a id="batch-multiple-prompts" 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="#batch-multiple-prompts"><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><!---->Gom nhiều prompt thành một lô<!----></span></h2><!--]--><!----> <p>Truyền một danh sách các prompt vào pipeline để tạo một hình ảnh cho mỗi prompt.</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> Đã sao chép</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">import</span> matplotlib.pyplot <span class="hljs-keyword">as</span> plt
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

pipeline = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
    dtype=torch.float16,
    device_map=<span class="hljs-string">&quot;cuda&quot;</span>  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
)

prompts = [
    <span class="hljs-string">&quot;Cinematic shot of a cozy coffee shop interior, warm pastel light streaming through a window where a cat rests. Shallow depth of field, glowing cups in soft focus, dreamy lofi-inspired mood, nostalgic tones, framed like a quiet film scene.&quot;</span>,
    <span class="hljs-string">&quot;Polaroid-style photograph of a cozy coffee shop interior, bathed in warm pastel light. A cat sits on the windowsill near steaming mugs. Soft, slightly faded tones and dreamy blur evoke nostalgia, a lofi mood, and the intimate, imperfect charm of instant film.&quot;</span>,
    <span class="hljs-string">&quot;Soft watercolor illustration of a cozy coffee shop interior, pastel washes of color filling the space. A cat rests peacefully on the windowsill as warm light glows through. Gentle brushstrokes create a dreamy, lofi-inspired atmosphere with whimsical textures and nostalgic calm.&quot;</span>,
    <span class="hljs-string">&quot;Isometric pixel-art illustration of a cozy coffee shop interior in detailed 8-bit style. Warm pastel light fills the space as a cat rests on the windowsill. Blocky furniture and tiny mugs add charm, low-res retro graphics enhance the nostalgic, lofi-inspired game aesthetic.&quot;</span>
]

images = pipeline(
    prompt=prompts,
).images

fig, axes = plt.subplots(<span class="hljs-number">2</span>, <span class="hljs-number">2</span>, figsize=(<span class="hljs-number">12</span>, <span class="hljs-number">12</span>))
axes = axes.flatten()

<span class="hljs-keyword">for</span> i, image <span class="hljs-keyword">in</span> <span class="hljs-built_in">enumerate</span>(images):
    axes[i].imshow(image)
    axes[i].set_title(<span class="hljs-string">f&quot;Image <span class="hljs-subst">{i+<span class="hljs-number">1</span>}</span>&quot;</span>)
    axes[i].axis(<span class="hljs-string">&#x27;off&#x27;</span>)

plt.tight_layout()
plt.show()<!----></pre></div><!----> <div class="flex justify-center"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/batch-inference.png"/></div> <!--[1--><h2 class="relative group"><a id="generate-several-images-per-prompt" 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="#generate-several-images-per-prompt"><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><!---->Tạo nhiều hình ảnh cho mỗi prompt<!----></span></h2><!--]--><!----> <p>Để tạo nhiều biến thể của một prompt, hãy sử dụng đối số <code>num_images_per_prompt</code>.</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> Đã sao chép</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">import</span> matplotlib.pyplot <span class="hljs-keyword">as</span> plt
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

pipeline = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
    dtype=torch.float16,
    device_map=<span class="hljs-string">&quot;cuda&quot;</span>  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
)

prompt=<span class="hljs-string">&quot;&quot;&quot;
Isometric pixel-art illustration of a cozy coffee shop interior in detailed 8-bit style. Warm pastel light fills the
space as a cat rests on the windowsill. Blocky furniture and tiny mugs add charm, low-res retro graphics enhance the
nostalgic, lofi-inspired game aesthetic.
&quot;&quot;&quot;</span>

images = pipeline(
    prompt=prompt,
    num_images_per_prompt=<span class="hljs-number">4</span>
).images

fig, axes = plt.subplots(<span class="hljs-number">2</span>, <span class="hljs-number">2</span>, figsize=(<span class="hljs-number">12</span>, <span class="hljs-number">12</span>))
axes = axes.flatten()

<span class="hljs-keyword">for</span> i, image <span class="hljs-keyword">in</span> <span class="hljs-built_in">enumerate</span>(images):
    axes[i].imshow(image)
    axes[i].set_title(<span class="hljs-string">f&quot;Image <span class="hljs-subst">{i+<span class="hljs-number">1</span>}</span>&quot;</span>)
    axes[i].axis(<span class="hljs-string">&#x27;off&#x27;</span>)

plt.tight_layout()
plt.show()<!----></pre></div><!----> <div class="flex justify-center"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/batch-inference-2.png"/></div> <!--[1--><h2 class="relative group"><a id="generate-several-images-for-several-prompts" 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="#generate-several-images-for-several-prompts"><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><!---->Tạo một vài hình ảnh cho một vài prompt<!----></span></h2><!--]--><!----> <p>Kết hợp cả hai cách tiếp cận để tạo ra các biến thể khác nhau của các prompt khác nhau. Ví dụ này tái sử dụng <code>pipeline</code> và <code>prompts</code> từ các ví dụ ở trên.</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> Đã sao chép</div><!----></button><!----></div> <pre class="language-py "><!---->images = pipeline(
    prompt=prompts,
    num_images_per_prompt=<span class="hljs-number">2</span>,
).images

fig, axes = plt.subplots(<span class="hljs-number">2</span>, <span class="hljs-number">4</span>, figsize=(<span class="hljs-number">12</span>, <span class="hljs-number">12</span>))
axes = axes.flatten()

<span class="hljs-keyword">for</span> i, image <span class="hljs-keyword">in</span> <span class="hljs-built_in">enumerate</span>(images):
    axes[i].imshow(image)
    axes[i].set_title(<span class="hljs-string">f&quot;Image <span class="hljs-subst">{i+<span class="hljs-number">1</span>}</span>&quot;</span>)
    axes[i].axis(<span class="hljs-string">&#x27;off&#x27;</span>)

plt.tight_layout()
plt.show()<!----></pre></div><!----> <div class="flex justify-center"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/batch-inference-3.png"/></div> <!--[1--><h2 class="relative group"><a id="generate-reproducible-batches" 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="#generate-reproducible-batches"><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><!---->Tạo các lô dữ liệu có thể tái lập<!----></span></h2><!--]--><!----> <p>Để một lô (batch) có thể tái lập được, hãy truyền cho pipeline một danh sách các đối tượng <a href="https://pytorch.org/docs/stable/generated/torch.Generator.html" rel="nofollow">Generator</a>, mỗi đối tượng tương ứng với một hình ảnh, mỗi đối tượng có một seed riêng.</p> <blockquote class="tip"><p>Tham khảo tài liệu <a href="./reusing_seeds">Reproducibility</a> để tìm hiểu thêm về các thuật toán tất định và đối tượng <code>Generator</code>.</p></blockquote> <p>Tạo một <code>Generator</code> cho mỗi hình ảnh bằng list comprehension. Nhân một list (<code>[generator] * 3</code>) sẽ cho bạn ba tham chiếu đến cùng một <code>Generator</code>, do đó các hình ảnh không nhận được các seed độc lập.</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> Đã sao chép</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-comment"># don&#x27;t do this</span>
generator = [torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">0</span>)] * <span class="hljs-number">3</span>  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span><!----></pre></div><!----> <p>Số lượng generator phải khớp với số lượng ảnh trong batch. Truyền <code>generator</code> vào pipeline.</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> Đã sao chép</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">import</span> matplotlib.pyplot <span class="hljs-keyword">as</span> plt
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

pipeline = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
    dtype=torch.float16,
    device_map=<span class="hljs-string">&quot;cuda&quot;</span>
)

prompts = [
    <span class="hljs-string">&quot;Cinematic shot of a cozy coffee shop interior, warm pastel light streaming through a window where a cat rests. Shallow depth of field, glowing cups in soft focus, dreamy lofi-inspired mood, nostalgic tones, framed like a quiet film scene.&quot;</span>,
    <span class="hljs-string">&quot;Polaroid-style photograph of a cozy coffee shop interior, bathed in warm pastel light. A cat sits on the windowsill near steaming mugs. Soft, slightly faded tones and dreamy blur evoke nostalgia, a lofi mood, and the intimate, imperfect charm of instant film.&quot;</span>,
    <span class="hljs-string">&quot;Soft watercolor illustration of a cozy coffee shop interior, pastel washes of color filling the space. A cat rests peacefully on the windowsill as warm light glows through. Gentle brushstrokes create a dreamy, lofi-inspired atmosphere with whimsical textures and nostalgic calm.&quot;</span>,
    <span class="hljs-string">&quot;Isometric pixel-art illustration of a cozy coffee shop interior in detailed 8-bit style. Warm pastel light fills the space as a cat rests on the windowsill. Blocky furniture and tiny mugs add charm, low-res retro graphics enhance the nostalgic, lofi-inspired game aesthetic.&quot;</span>
]
generator = [torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(i) <span class="hljs-keyword">for</span> i <span class="hljs-keyword">in</span> <span class="hljs-built_in">range</span>(<span class="hljs-built_in">len</span>(prompts))]  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

images = pipeline(
    prompt=prompts,
    generator=generator
).images

fig, axes = plt.subplots(<span class="hljs-number">2</span>, <span class="hljs-number">2</span>, figsize=(<span class="hljs-number">12</span>, <span class="hljs-number">12</span>))
axes = axes.flatten()

<span class="hljs-keyword">for</span> i, image <span class="hljs-keyword">in</span> <span class="hljs-built_in">enumerate</span>(images):
    axes[i].imshow(image)
    axes[i].set_title(<span class="hljs-string">f&quot;Image <span class="hljs-subst">{i+<span class="hljs-number">1</span>}</span>&quot;</span>)
    axes[i].axis(<span class="hljs-string">&#x27;off&#x27;</span>)

plt.tight_layout()
plt.show()<!----></pre></div><!----> <p>Mỗi hình ảnh được gắn với một seed, vì vậy bạn có thể chọn một hình ảnh mà bạn thích và tạo lại nó với cùng một seed trong khi bạn tinh chỉnh prompt của nó.</p> <!--[1--><h2 class="relative group"><a id="serve-concurrent-requests" 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="#serve-concurrent-requests"><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><!---->Xử lý các yêu cầu đồng thời<!----></span></h2><!--]--><!----> <p>Máy chủ cũng tăng thông lượng bằng cách xử lý các yêu cầu từ nhiều người dùng cùng lúc. Nạp pipeline một lần và chia sẻ nó giữa các yêu cầu để các mô hình chỉ nằm trong bộ nhớ GPU một lần. Scheduler không an toàn cho đa luồng, vì vậy hãy cấp cho mỗi yêu cầu một scheduler riêng bằng <a href="/docs/diffusers/main/vi/api/pipelines/overview#diffusers.DiffusionPipeline.from_pipe">from_pipe()</a>. Chia sẻ một scheduler duy nhất giữa các luồng sẽ gây ra các lỗi như <code>IndexError: index 21 is out of bounds for dimension 0 with size 21</code>.</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> Đã sao chép</div><!----></button><!----></div> <pre class="language-py "><!---->scheduler = shared_pipeline.scheduler.from_config(shared_pipeline.scheduler.config)
pipeline = StableDiffusion3Pipeline.from_pipe(shared_pipeline, scheduler=scheduler)<!----></pre></div><!----> <p>Đối với các máy chủ hoàn chỉnh được xây dựng bằng <a href="https://fastapi.tiangolo.com/" rel="nofollow">FastAPI</a>, hãy xem ví dụ <a href="https://github.com/huggingface/diffusers/tree/main/examples/server" rel="nofollow">server</a> và ví dụ <a href="https://github.com/huggingface/diffusers/tree/main/examples/server-async" rel="nofollow">server-async</a>. Ví dụ server-async cũng sao chép trạng thái của bộ tạo số ngẫu nhiên cho mỗi yêu cầu và đảm bảo các tokenizer an toàn khi chia sẻ giữa các luồng.</p> <!--[1--><h2 class="relative group"><a id="next-steps" 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="#next-steps"><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><!---->Các bước tiếp theo<!----></span></h2><!--]--><!----> <ul><li>Khi một batch không vừa trên một GPU, hoặc bạn có nhiều prompt hơn số lượng mà một GPU có thể xử lý, hãy phân tán các prompt trên nhiều GPU bằng <a href="../optimization/data_parallelism">Data parallelism</a>.</li></ul> <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/using-diffusers/batched_inference.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_1qsdbwf = {
						base: "/docs/diffusers/main/en",
						assets: "/docs/diffusers/main/en"
					};
					const element = document.currentScript.parentElement;
					Promise.all([
						import("/docs/diffusers/main/en/_app/immutable/entry/start.0S0_iEoG.js"),
						import("/docs/diffusers/main/en/_app/immutable/entry/app.CPdOxCbw.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 386],
							data: [null,null],
							form: null,
							error: null
						});
					});
				}
			</script>
		
Xet Storage Details

Size:: 34.7 kB
Xet hash:: 6f5a3894173d4babe789285253f01754b1a1fba3287cfb26c168134d9f2886ef
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.