Buckets:

hf-doc-build
/

doc

Files

xet

hf-doc-build/doc / diffusers /main /en /modular_diffusers /quickstart.html

HuggingFaceDocBuilder

about 7 hours ago

download

raw

54.8 kB

	<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Quickstart","local":"quickstart","sections":[{"title":"Run a pipeline","local":"run-a-pipeline","sections":[],"depth":2},{"title":"Understand the structure","local":"understand-the-structure","sections":[{"title":"Workflows","local":"workflows","sections":[],"depth":3},{"title":"Sub-blocks","local":"sub-blocks","sections":[],"depth":3}],"depth":2},{"title":"Compose new workflows","local":"compose-new-workflows","sections":[],"depth":2},{"title":"Next steps","local":"next-steps","sections":[],"depth":2}],"depth":1}">
	<link href="/docs/diffusers/main/en/_app/immutable/assets/0.e3b0c442.css" rel="modulepreload">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/entry/start.55138708.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/scheduler.53228c21.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/singletons.59fc0d32.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/index.e93d0901.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/paths.4f58d9a7.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/entry/app.9b1131ae.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/preload-helper.5c375679.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/index.100fac89.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/nodes/0.eac8817e.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/each.e59479a4.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/nodes/264.ca3e6b36.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/CopyLLMTxtMenu.37908ba4.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/globals.7f7f1b26.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/IconCopy.38cf8f56.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/MermaidChart.svelte_svelte_type_style_lang.0a9663c0.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/CodeBlock.d30a6509.js">
	<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/HfOption.fad27e59.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{"title":"Quickstart","local":"quickstart","sections":[{"title":"Run a pipeline","local":"run-a-pipeline","sections":[],"depth":2},{"title":"Understand the structure","local":"understand-the-structure","sections":[{"title":"Workflows","local":"workflows","sections":[],"depth":3},{"title":"Sub-blocks","local":"sub-blocks","sections":[],"depth":3}],"depth":2},{"title":"Compose new workflows","local":"compose-new-workflows","sections":[],"depth":2},{"title":"Next steps","local":"next-steps","sections":[],"depth":2}],"depth":1}"><!-- HEAD_svelte-u9bgzb_END --> <p></p> <div class="items-center shrink-0 min-w-[100px] max-sm:min-w-[50px] justify-end ml-auto flex" style="float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"><div 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>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> </div> <h1 class="relative group"><a id="quickstart" 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="#quickstart"><span><svg class="" 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>Quickstart</span></h1> <p data-svelte-h="svelte-16nb4am">Modular Diffusers is a framework for quickly building flexible and customizable pipelines. These pipelines can go beyond what standard <code>DiffusionPipeline</code>s can do. At the core of Modular Diffusers are <a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline_blocks#diffusers.ModularPipelineBlocks">ModularPipelineBlocks</a> that can be combined with other blocks to adapt to new workflows. The blocks are converted into a <a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline#diffusers.ModularPipeline">ModularPipeline</a>, a friendly user-facing interface for running generation tasks.</p> <p data-svelte-h="svelte-1wfnmh6">This guide shows you how to run a modular pipeline, understand its structure, and customize it by modifying the blocks that compose it.</p> <h2 class="relative group"><a id="run-a-pipeline" 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="#run-a-pipeline"><span><svg class="" 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>Run a pipeline</span></h2> <p data-svelte-h="svelte-frlpr1"><a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline#diffusers.ModularPipeline">ModularPipeline</a> is the main interface for loading, running, and managing modular pipelines.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">import</span> torch
	<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> ModularPipeline, ComponentsManager

	<span class="hljs-comment"># Use ComponentsManager to enable auto CPU offloading for memory efficiency</span>
	manager = ComponentsManager()
	manager.enable_auto_cpu_offload(device=<span class="hljs-string">"cuda:0"</span>)

	pipe = ModularPipeline.from_pretrained(<span class="hljs-string">"Qwen/Qwen-Image"</span>, components_manager=manager)
	pipe.load_components(torch_dtype=torch.bfloat16)

	image = pipe(
	prompt=<span class="hljs-string">"cat wizard with red hat, gandalf, lord of the rings, detailed, fantasy, cute, adorable, Pixar, Disney"</span>,
	).images[<span class="hljs-number">0</span>]
	image<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1mcr55h"><a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline#diffusers.ModularPipeline.from_pretrained">from_pretrained()</a> uses lazy loading - it reads the configuration to learn where to load each component from, but doesn’t actually load the model weights until you call <a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline#diffusers.ModularPipeline.load_components">load_components()</a>. This gives you control over when and how components are loaded.</p> <blockquote class="tip" data-svelte-h="svelte-s8c2id"><p><code>ComponentsManager</code> with <code>enable_auto_cpu_offload</code> automatically moves models between CPU and GPU as needed, reducing memory usage for large models like Qwen-Image. Learn more in the <a href="./components_manager">ComponentsManager</a> guide.</p> <p>If you don’t need offloading, remove the <code>components_manager</code> argument and move the pipeline to your device manually with <code>to("cuda")</code>.</p></blockquote> <p data-svelte-h="svelte-bpx8o">Learn more about creating and loading pipelines in the <a href="https://huggingface.co/docs/diffusers/modular_diffusers/modular_pipeline#creating-a-pipeline" rel="nofollow">Creating a pipeline</a> and <a href="https://huggingface.co/docs/diffusers/modular_diffusers/modular_pipeline#loading-components" rel="nofollow">Loading components</a> guides.</p> <h2 class="relative group"><a id="understand-the-structure" 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="#understand-the-structure"><span><svg class="" 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>Understand the structure</span></h2> <p data-svelte-h="svelte-1pe3e8n">A <a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline#diffusers.ModularPipeline">ModularPipeline</a> has two parts: a <strong>definition</strong> (the blocks) and a <strong>state</strong> (the loaded components and configs).</p> <p data-svelte-h="svelte-1dhi4po">Print the pipeline to see its state — the components and their loading status and configuration.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-built_in">print</span>(pipe)<!-- HTML_TAG_END --></pre></div> <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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-symbol">QwenImageModularPipeline</span> {
	<span class="hljs-string">"_blocks_class_name"</span>: <span class="hljs-string">"QwenImageAutoBlocks"</span>,
	<span class="hljs-string">"_class_name"</span>: <span class="hljs-string">"QwenImageModularPipeline"</span>,
	<span class="hljs-string">"_diffusers_version"</span>: <span class="hljs-string">"0.37.0.dev0"</span>,
	<span class="hljs-string">"transformer"</span>: [
	<span class="hljs-string">"diffusers"</span>,
	<span class="hljs-string">"QwenImageTransformer2DModel"</span>,
	{
	<span class="hljs-string">"pretrained_model_name_or_path"</span>: <span class="hljs-string">"Qwen/Qwen-Image"</span>,
	<span class="hljs-string">"revision"</span>: null,
	<span class="hljs-string">"subfolder"</span>: <span class="hljs-string">"transformer"</span>,
	<span class="hljs-string">"type_hint"</span>: [
	<span class="hljs-string">"diffusers"</span>,
	<span class="hljs-string">"QwenImageTransformer2DModel"</span>
	],
	<span class="hljs-string">"variant"</span>: null
	}
	],
	...
	}<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-fhlizn">Access the definition through <code>pipe.blocks</code> — this is the <a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline_blocks#diffusers.ModularPipelineBlocks">ModularPipelineBlocks</a> that defines the pipeline’s workflows, inputs, outputs, and computation logic.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-built_in">print</span>(pipe.blocks)<!-- HTML_TAG_END --></pre></div> <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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->QwenImageAutoBlocks(
	Class: SequentialPipelineBlocks

	Description: Auto Modular pipeline <span class="hljs-keyword">for</span> <span class="hljs-keyword">text</span>-to-<span class="hljs-keyword">image</span>, <span class="hljs-keyword">image</span>-to-<span class="hljs-keyword">image</span>, inpainting, and controlnet tasks using QwenImage.

	Supported workflows:
	- <span class="hljs-string">`text2image`</span>: <span class="hljs-keyword">requires</span> <span class="hljs-string">`prompt`</span>
	- <span class="hljs-string">`image2image`</span>: <span class="hljs-keyword">requires</span> <span class="hljs-string">`prompt`</span>, <span class="hljs-string">`image`</span>
	- <span class="hljs-string">`inpainting`</span>: <span class="hljs-keyword">requires</span> <span class="hljs-string">`prompt`</span>, <span class="hljs-string">`mask_image`</span>, <span class="hljs-string">`image`</span>
	- <span class="hljs-string">`controlnet_text2image`</span>: <span class="hljs-keyword">requires</span> <span class="hljs-string">`prompt`</span>, <span class="hljs-string">`control_image`</span>
	...

	Components:
	text_encoder (<span class="hljs-string">`Qwen2_5_VLForConditionalGeneration`</span>)
	vae (<span class="hljs-string">`AutoencoderKLQwenImage`</span>)
	transformer (<span class="hljs-string">`QwenImageTransformer2DModel`</span>)
	...

	Sub-Blocks:
	[<span class="hljs-number">0</span>] text_encoder (QwenImageAutoTextEncoderStep)
	[<span class="hljs-number">1</span>] vae_encoder (QwenImageAutoVaeEncoderStep)
	[<span class="hljs-number">2</span>] controlnet_vae_encoder (QwenImageOptionalControlNetVaeEncoderStep)
	[<span class="hljs-number">3</span>] denoise (QwenImageAutoCoreDenoiseStep)
	[<span class="hljs-number">4</span>] decode (QwenImageAutoDecodeStep)
	)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-e3gjkj">The output returns:</p> <ul data-svelte-h="svelte-nqrxsq"><li>The supported workflows (text2image, image2image, inpainting, etc.)</li> <li>The Sub-Blocks it’s composed of (text_encoder, vae_encoder, denoise, decode)</li></ul> <h3 class="relative group"><a id="workflows" 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="#workflows"><span><svg class="" 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>Workflows</span></h3> <p data-svelte-h="svelte-vcci71">This pipeline supports multiple workflows and adapts its behavior based on the inputs you provide. For example, if you pass <code>image</code> to the pipeline, it runs an image-to-image workflow instead of text-to-image. Learn more about how this works under the hood in the <a href="https://huggingface.co/docs/diffusers/modular_diffusers/auto_pipeline_blocks" rel="nofollow">AutoPipelineBlocks</a> guide.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

	input_image = load_image(<span class="hljs-string">"https://github.com/Trgtuan10/Image_storage/blob/main/cute_cat.png?raw=true"</span>)

	image = pipe(
	prompt=<span class="hljs-string">"cat wizard with red hat, gandalf, lord of the rings, detailed, fantasy, cute, adorable, Pixar, Disney"</span>,
	image=input_image,
	).images[<span class="hljs-number">0</span>]<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-xbf9lq">Use <code>get_workflow()</code> to extract the blocks for a specific workflow. Pass the workflow name (e.g., <code>"image2image"</code>, <code>"inpainting"</code>, <code>"controlnet_text2image"</code>) to get only the blocks relevant to that workflow. This is useful when you want to customize or debug a specific workflow. You can check <code>pipe.blocks.available_workflows</code> to see all available workflows.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->img2img_blocks = pipe.blocks.get_workflow(<span class="hljs-string">"image2image"</span>)<!-- HTML_TAG_END --></pre></div> <h3 class="relative group"><a id="sub-blocks" 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="#sub-blocks"><span><svg class="" 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>Sub-blocks</span></h3> <p data-svelte-h="svelte-1ewhq0u">Blocks can contain other blocks. <code>pipe.blocks</code> gives you the top-level block definition (here, <code>QwenImageAutoBlocks</code>), while <code>sub_blocks</code> lets you access the smaller blocks inside it.</p> <p data-svelte-h="svelte-15ryot1"><code>QwenImageAutoBlocks</code> is composed of: <code>text_encoder</code>, <code>vae_encoder</code>, <code>controlnet_vae_encoder</code>, <code>denoise</code>, and <code>decode</code>.</p> <p data-svelte-h="svelte-1munkh1">These sub-blocks run one after another and data flows linearly from one block to the next — each block’s <code>intermediate_outputs</code> become available as <code>inputs</code> to the next block. This is how <a href="./sequential_pipeline_blocks"><code>SequentialPipelineBlocks</code></a> work.</p> <p data-svelte-h="svelte-i0k50u">You can access them through the <code>sub_blocks</code> property. The <code>doc</code> property is useful for seeing the full documentation of any block, including its inputs, outputs, and components.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->vae_encoder_block = pipe.blocks.sub_blocks[<span class="hljs-string">"vae_encoder"</span>]
	<span class="hljs-built_in">print</span>(vae_encoder_block.doc)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-ax5nee">This block can be converted to a pipeline so that it can run on its own with <a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline_blocks#diffusers.ModularPipelineBlocks.init_pipeline">init_pipeline()</a>.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->vae_encoder_pipe = vae_encoder_block.init_pipeline()

	<span class="hljs-comment"># Reuse the VAE we already loaded, we can reuse it with update_components() method</span>
	vae_encoder_pipe.update_components(vae=pipe.vae)

	<span class="hljs-comment"># Run just this block</span>
	image_latents = vae_encoder_pipe(image=input_image).image_latents
	<span class="hljs-built_in">print</span>(image_latents.shape)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-l37vyv">It reuses the VAE from our original pipeline instead of reloading it, keeping memory usage efficient. Learn more in the <a href="https://huggingface.co/docs/diffusers/modular_diffusers/modular_pipeline#loading-components" rel="nofollow">Loading components</a> guide.</p> <p data-svelte-h="svelte-1a0h615">Since blocks are composable, you can modify the pipeline’s definition by adding, removing, or swapping blocks to create new workflows. In the next section, we’ll add a canny edge detection block to a ControlNet pipeline, so you can pass a regular image instead of a pre-processed canny edge map.</p> <h2 class="relative group"><a id="compose-new-workflows" 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="#compose-new-workflows"><span><svg class="" 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>Compose new workflows</span></h2> <p data-svelte-h="svelte-1xhlgp">Let’s add a canny edge detection block to a ControlNet pipeline. First, load a pre-built canny block from the Hub (see <a href="https://huggingface.co/docs/diffusers/modular_diffusers/custom_blocks" rel="nofollow">Building Custom Blocks</a> to create your own).</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">from</span> diffusers.modular_pipelines <span class="hljs-keyword">import</span> ModularPipelineBlocks

	<span class="hljs-comment"># Load a canny block from the Hub</span>
	canny_block = ModularPipelineBlocks.from_pretrained(
	<span class="hljs-string">"diffusers-internal-dev/canny-filtering"</span>,
	trust_remote_code=<span class="hljs-literal">True</span>,
	)

	<span class="hljs-built_in">print</span>(canny_block.doc)<!-- HTML_TAG_END --></pre></div> <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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->class CannyBlock

	Inputs:
	image (`Union[Image, ndarray]`):
	Image <span class="hljs-keyword">to</span> compute canny <span class="hljs-built_in">filter</span> <span class="hljs-keyword">on</span>
	low_threshold (`<span class="hljs-keyword">int</span>`, optional, defaults <span class="hljs-keyword">to</span> <span class="hljs-number">50</span>):
	Low threshold <span class="hljs-keyword">for</span> the canny <span class="hljs-built_in">filter</span>.
	high_threshold (`<span class="hljs-keyword">int</span>`, optional, defaults <span class="hljs-keyword">to</span> <span class="hljs-number">200</span>):
	High threshold <span class="hljs-keyword">for</span> the canny <span class="hljs-built_in">filter</span>.
	...

	Outputs:
	control_image (`PIL.Image`):
	Canny <span class="hljs-keyword">map</span> <span class="hljs-keyword">for</span> <span class="hljs-built_in">input</span> image<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-xaa03y">Use <code>get_workflow</code> to extract the ControlNet workflow from <code>QwenImageAutoBlocks</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-comment"># Get the controlnet workflow that we want to work with</span>
	blocks = pipe.blocks.get_workflow(<span class="hljs-string">"controlnet_text2image"</span>)
	<span class="hljs-built_in">print</span>(blocks.doc)<!-- HTML_TAG_END --></pre></div> <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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->class <span class="hljs-symbol">SequentialPipelineBlocks</span>

	<span class="hljs-symbol">Inputs</span>:
	prompt (<span class="hljs-string">`str`</span>):
	<span class="hljs-symbol">The</span> prompt or prompts to guide image generation.
	control_image (<span class="hljs-string">`Image`</span>):
	<span class="hljs-symbol">Control</span> image for <span class="hljs-symbol">ControlNet</span> conditioning.
	...<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1wl8tkn">The extracted workflow is a <a href="./sequential_pipeline_blocks"><code>SequentialPipelineBlocks</code></a> and it currently requires <code>control_image</code> as input. Insert the canny block at the beginning so the pipeline accepts a regular image instead.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-comment"># Insert canny at the beginning</span>
	blocks.sub_blocks.insert(<span class="hljs-string">"canny"</span>, canny_block, <span class="hljs-number">0</span>)

	<span class="hljs-comment"># Check the updated structure: CannyBlock is now listed as first sub-block</span>
	<span class="hljs-built_in">print</span>(blocks)
	<span class="hljs-comment"># Check the updated doc</span>
	<span class="hljs-built_in">print</span>(blocks.doc)<!-- HTML_TAG_END --></pre></div> <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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->class SequentialPipelineBlocks

	Inputs:
	image (<span class="hljs-string">`Union[Image, ndarray]`</span>):
	Image <span class="hljs-keyword">to</span> compute canny filter <span class="hljs-keyword">on</span>
	low_threshold (<span class="hljs-string">`int`</span>, optional, defaults <span class="hljs-keyword">to</span> <span class="hljs-number">50</span>):
	Low threshold for the canny filter.
	high_threshold (<span class="hljs-string">`int`</span>, optional, defaults <span class="hljs-keyword">to</span> <span class="hljs-number">200</span>):
	High threshold for the canny filter.
	prompt (<span class="hljs-string">`str`</span>):
	The prompt <span class="hljs-literal">or</span> prompts <span class="hljs-keyword">to</span> guide image generation.
	<span class="hljs-params">...</span><!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1u9dq8d">Now the pipeline takes <code>image</code> as input instead of <code>control_image</code>. Because blocks in a sequence share data automatically, the canny block’s output (<code>control_image</code>) flows to the denoise block that needs it, and the canny block’s input (<code>image</code>) becomes a pipeline input since no earlier block provides it.</p> <p data-svelte-h="svelte-1vo03c8">Create a pipeline from the modified blocks and load a ControlNet model. The ControlNet isn’t part of the original model repository, so load it separately and add it with <a href="/docs/diffusers/main/en/api/modular_diffusers/pipeline#diffusers.ModularPipeline.update_components">update_components()</a>.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START -->pipeline = blocks.init_pipeline(<span class="hljs-string">"Qwen/Qwen-Image"</span>, components_manager=manager)

	pipeline.load_components(torch_dtype=torch.bfloat16)

	<span class="hljs-comment"># Load the ControlNet model</span>
	controlnet_spec = pipeline.get_component_spec(<span class="hljs-string">"controlnet"</span>)
	controlnet_spec.pretrained_model_name_or_path = <span class="hljs-string">"InstantX/Qwen-Image-ControlNet-Union"</span>
	controlnet = controlnet_spec.load(torch_dtype=torch.bfloat16)
	pipeline.update_components(controlnet=controlnet)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-41vcz">Now run the pipeline - the canny block preprocesses the image for ControlNet.</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 class="" 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> Copied</div></button></div> <pre class=""><!-- HTML_TAG_START --><span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

	prompt = <span class="hljs-string">"cat wizard with red hat, gandalf, lord of the rings, detailed, fantasy, cute, adorable, Pixar, Disney"</span>
	image = load_image(<span class="hljs-string">"https://github.com/Trgtuan10/Image_storage/blob/main/cute_cat.png?raw=true"</span>)

	output = pipeline(
	prompt=prompt,
	image=image,
	).images[<span class="hljs-number">0</span>]
	output<!-- HTML_TAG_END --></pre></div> <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 class="" 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>Next steps</span></h2> <div class="flex space-x-2 items-center my-1.5 mr-8 h-7 !pl-0 -mx-3 md:mx-0"><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd border-gray-800 bg-black dark:bg-gray-700 text-white">Learn the basics </div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">Build custom blocks </div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">Share components </div><div class="flex items-center border rounded-lg px-1.5 py-1 leading-none select-none text-smd text-gray-500 cursor-pointer opacity-90 hover:text-gray-700 dark:hover:text-gray-200 hover:shadow-sm">Visual interface </div></div> <div class="language-select"><p data-svelte-h="svelte-f5cfxq">Understand the core building blocks of Modular Diffusers:</p> <ul data-svelte-h="svelte-7ovs2f"><li><a href="./pipeline_block">ModularPipelineBlocks</a>: The basic unit for defining a step in a pipeline.</li> <li><a href="./sequential_pipeline_blocks">SequentialPipelineBlocks</a>: Chain blocks to run in sequence.</li> <li><a href="./auto_pipeline_blocks">AutoPipelineBlocks</a>: Create pipelines that support multiple workflows.</li> <li><a href="./modular_diffusers_states">States</a>: How data is shared between blocks.</li></ul> </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/modular_diffusers/quickstart.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 data-svelte-h="svelte-zjs2n5"><span class="underline">Update</span> on GitHub</span></a> <p></p>

	<script>
	{
	__sveltekit_19i7tg0 = {
	assets: "/docs/diffusers/main/en",
	base: "/docs/diffusers/main/en",
	env: {}
	};

	const element = document.currentScript.parentElement;

	const data = [null,null];

	Promise.all([
	import("/docs/diffusers/main/en/_app/immutable/entry/start.55138708.js"),
	import("/docs/diffusers/main/en/_app/immutable/entry/app.9b1131ae.js")
	]).then(([kit, app]) => {
	kit.start(app, element, {
	node_ids: [0, 264],
	data,
	form: null,
	error: null
	});
	});
	}
	</script>

Xet Storage Details

Size:: 54.8 kB
Xet hash:: a282e4bc802ac3ff785befdcf971626d2d022c3ca6148145cbde4366c8bcffd1

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