Buckets:

hf-doc-build
/

doc-dev

Files
xet
hf-doc-build/doc-dev / diffusers /pr_15230 /en /using-diffusers /ip_adapter.html
HuggingFaceDocBuilder
about 24 hours ago
download
raw
90.4 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;IP-Adapter&quot;,&quot;local&quot;:&quot;ip-adapter&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Checkpoint variants&quot;,&quot;local&quot;:&quot;checkpoint-variants&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Image embeddings&quot;,&quot;local&quot;:&quot;image-embeddings&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Masking&quot;,&quot;local&quot;:&quot;masking&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Recipes&quot;,&quot;local&quot;:&quot;recipes&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Multiple IP-Adapters&quot;,&quot;local&quot;:&quot;multiple-ip-adapters&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Structural control&quot;,&quot;local&quot;:&quot;structural-control&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Style and layout control&quot;,&quot;local&quot;:&quot;style-and-layout-control&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Instant generation&quot;,&quot;local&quot;:&quot;instant-generation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/>
		<link href="/docs/diffusers/main/en/_app/immutable/entry/start.vZc-dOUR.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/Bafuik5c.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.B6c3Ml05.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.BQlx092V.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/392.CkZ-gPeA.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/B4AMoTAu.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/B_Tpb3TG.js" rel="modulepreload">
		<!--1p5pew9--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;IP-Adapter&quot;,&quot;local&quot;:&quot;ip-adapter&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Checkpoint variants&quot;,&quot;local&quot;:&quot;checkpoint-variants&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Image embeddings&quot;,&quot;local&quot;:&quot;image-embeddings&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Masking&quot;,&quot;local&quot;:&quot;masking&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Recipes&quot;,&quot;local&quot;:&quot;recipes&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Multiple IP-Adapters&quot;,&quot;local&quot;:&quot;multiple-ip-adapters&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Structural control&quot;,&quot;local&quot;:&quot;structural-control&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Style and layout control&quot;,&quot;local&quot;:&quot;style-and-layout-control&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Instant generation&quot;,&quot;local&quot;:&quot;instant-generation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&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="ip-adapter" 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="#ip-adapter"><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><!---->IP-Adapter<!----></span></h1><!--]--><!----> <p><a href="https://huggingface.co/papers/2308.06721" rel="nofollow">IP-Adapter</a> steers a pretrained diffusion model with a reference image while you keep a text prompt. It freezes the base model and adds a small set of image cross-attention layers, which makes it practical for matching a subject or style from a photo without fine-tuning.</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> Copied</div><!----></button><!----></div> <pre class="language-text "><!---->text prompt                         reference image
    |                                     |
 text encoder                        image encoder
    |                                     |
    v                                     v
 text cross-attn  (frozen)          IP-Adapter cross-attn
    \                                     /
     \                                   /
      +--------&gt;  denoiser  &lt;-----------+
                  (frozen base)<!----></pre></div><!----> <p>IP-Adapter checkpoints are typically ~100MB because they store adapter weights, not a full model. Load a base pipeline first, then load the adapter with <a href="/docs/diffusers/pr_15230/en/api/loaders/ip_adapter#diffusers.loaders.IPAdapterMixin.load_ip_adapter">load_ip_adapter()</a>.</p> <blockquote class="tip"><p>IP-Adapters are available to many models such as <a href="../api/pipelines/flux#ip-adapter">Flux</a> and <a href="../api/pipelines/stable_diffusion/stable_diffusion_3">Stable Diffusion 3</a>, and more. The examples in this guide use Stable Diffusion and Stable Diffusion XL.</p></blockquote> <p>Use <a href="/docs/diffusers/pr_15230/en/api/loaders/ip_adapter#diffusers.loaders.IPAdapterMixin.set_ip_adapter_scale">set_ip_adapter_scale()</a> to control how strongly the IP-Adapter steers generation. <code>1.0</code> applies the adapter at full strength; <code>0.5</code> usually balances text and image prompts.</p> <p>The examples below show IP-Adapter on common tasks.</p> <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">text-to-image</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">image-to-image</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">inpainting</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">video</div><!--]--></div> <div class="language-select"><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 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="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForText2Image
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

pipeline = AutoPipelineForText2Image.from_pretrained(
  <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
  dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter_sdxl.bin&quot;</span>
)
pipeline.set_ip_adapter_scale(<span class="hljs-number">0.8</span>)<!----></pre></div><!----> <p>Pass an image to <code>ip_adapter_image</code> along with a text prompt to generate an image.</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_diner.png&quot;</span>)
pipeline(
    prompt=<span class="hljs-string">&quot;a polar bear sitting in a chair drinking a milkshake&quot;</span>,
    ip_adapter_image=image,
    negative_prompt=<span class="hljs-string">&quot;deformed, ugly, wrong proportion, low res, bad anatomy, worst quality, low quality&quot;</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_diner.png" width="400" alt="IP-Adapter image"/> <figcaption style="text-align: center;">IP-Adapter image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_diner_2.png" width="400" alt="generated image"/> <figcaption style="text-align: center;">generated image</figcaption></figure></div><!----></div><!----> <div hidden=""><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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForImage2Image
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

pipeline = AutoPipelineForImage2Image.from_pretrained(
  <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
  dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter_sdxl.bin&quot;</span>
)
pipeline.set_ip_adapter_scale(<span class="hljs-number">0.8</span>)

image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_bear_1.png&quot;</span>)
ip_image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_gummy.png&quot;</span>)
pipeline(
    prompt=<span class="hljs-string">&quot;best quality, high quality&quot;</span>,
    image=image,
    ip_adapter_image=ip_image,
    strength=<span class="hljs-number">0.5</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_bear_1.png" width="300" alt="input image"/> <figcaption style="text-align: center;">input image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_gummy.png" width="300" alt="IP-Adapter image"/> <figcaption style="text-align: center;">IP-Adapter image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_bear_3.png" width="300" alt="generated image"/> <figcaption style="text-align: center;">generated image</figcaption></figure></div><!----></div><!----> <div hidden=""><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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForInpainting
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

pipeline = AutoPipelineForInpainting.from_pretrained(
  <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
  dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter_sdxl.bin&quot;</span>
)
pipeline.set_ip_adapter_scale(<span class="hljs-number">0.6</span>)

mask_image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_mask.png&quot;</span>)
image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_bear_1.png&quot;</span>)
ip_image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_gummy.png&quot;</span>)
pipeline(
    prompt=<span class="hljs-string">&quot;a cute gummy bear waving&quot;</span>,
    image=image,
    mask_image=mask_image,
    ip_adapter_image=ip_image,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_bear_1.png" width="300" alt="input image"/> <figcaption style="text-align: center;">input image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_gummy.png" width="300" alt="IP-Adapter image"/> <figcaption style="text-align: center;">IP-Adapter image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_inpaint.png" width="300" alt="generated image"/> <figcaption style="text-align: center;">generated image</figcaption></figure></div><!----></div><!----> <div hidden=""><p>The <a href="/docs/diffusers/pr_15230/en/api/pipelines/overview#diffusers.DiffusionPipeline.enable_model_cpu_offload">enable_model_cpu_offload()</a> method is useful for reducing memory and it should be enabled after the IP-Adapter is loaded. Otherwise, the IP-Adapter’s image encoder is also offloaded to the CPU and returns an error.</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AnimateDiffPipeline, DDIMScheduler, MotionAdapter
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> export_to_gif
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

adapter = MotionAdapter.from_pretrained(
  <span class="hljs-string">&quot;guoyww/animatediff-motion-adapter-v1-5-2&quot;</span>,
  dtype=torch.float16
)
pipeline = AnimateDiffPipeline.from_pretrained(
  <span class="hljs-string">&quot;emilianJR/epiCRealism&quot;</span>,
  motion_adapter=adapter,
  dtype=torch.float16
)
scheduler = DDIMScheduler.from_pretrained(
    <span class="hljs-string">&quot;emilianJR/epiCRealism&quot;</span>,
    subfolder=<span class="hljs-string">&quot;scheduler&quot;</span>,
    clip_sample=<span class="hljs-literal">False</span>,
    timestep_spacing=<span class="hljs-string">&quot;linspace&quot;</span>,
    beta_schedule=<span class="hljs-string">&quot;linear&quot;</span>,
    steps_offset=<span class="hljs-number">1</span>,
)
pipeline.scheduler = scheduler
pipeline.vae.enable_slicing()
pipeline.load_ip_adapter(<span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>, subfolder=<span class="hljs-string">&quot;models&quot;</span>, weight_name=<span class="hljs-string">&quot;ip-adapter_sd15.bin&quot;</span>)
pipeline.enable_model_cpu_offload()

ip_adapter_image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_inpaint.png&quot;</span>)
pipeline(
    prompt=<span class="hljs-string">&quot;A cute gummy bear waving&quot;</span>,
    negative_prompt=<span class="hljs-string">&quot;bad quality, worse quality, low resolution&quot;</span>,
    ip_adapter_image=ip_adapter_image,
    num_frames=<span class="hljs-number">16</span>,
    guidance_scale=<span class="hljs-number">7.5</span>,
    num_inference_steps=<span class="hljs-number">50</span>,
).frames[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_inpaint.png" width="400" alt="IP-Adapter image"/> <figcaption style="text-align: center;">IP-Adapter image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/gummy_bear.gif" width="400" alt="generated video"/> <figcaption style="text-align: center;">generated video</figcaption></figure></div><!----></div><!----><!----></div><!----> <!--[1--><h2 class="relative group"><a id="checkpoint-variants" 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="#checkpoint-variants"><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><!---->Checkpoint variants<!----></span></h2><!--]--><!----> <p>Load Plus when detail from the reference image matters most. Load FaceID when you need InsightFace identity embeddings rather than CLIP image embeddings.</p> <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">IP-Adapter Plus</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">IP-Adapter FaceID</div><!--]--></div> <div class="language-select"><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 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="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> CLIPVisionModelWithProjection
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForText2Image

image_encoder = CLIPVisionModelWithProjection.from_pretrained(
    <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
    subfolder=<span class="hljs-string">&quot;models/image_encoder&quot;</span>,
    dtype=torch.float16
)

pipeline = AutoPipelineForText2Image.from_pretrained(
    <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
    image_encoder=image_encoder,
    dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter-plus_sdxl_vit-h.safetensors&quot;</span>
)<!----></pre></div><!----></div><!----> <div hidden=""><p>FaceID checkpoints use face embeddings from <a href="https://github.com/deepinsight/insightface" rel="nofollow">InsightFace</a> instead of CLIP image embeddings. Use the <a href="/docs/diffusers/pr_15230/en/api/schedulers/ddim#diffusers.DDIMScheduler">DDIMScheduler</a> or <a href="/docs/diffusers/pr_15230/en/api/schedulers/euler#diffusers.EulerDiscreteScheduler">EulerDiscreteScheduler</a> for FaceID models. Extract the face embeddings and pass them as a list of tensors to <code>ip_adapter_image_embeds</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-comment"># pip install insightface</span>
<span class="hljs-keyword">import</span> cv2
<span class="hljs-keyword">import</span> numpy <span class="hljs-keyword">as</span> np
<span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> StableDiffusionPipeline, DDIMScheduler
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image
<span class="hljs-keyword">from</span> insightface.app <span class="hljs-keyword">import</span> FaceAnalysis

pipeline = StableDiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>,
    dtype=torch.float16,
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline.scheduler = DDIMScheduler.from_config(pipeline.scheduler.config)
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter-FaceID&quot;</span>,
  subfolder=<span class="hljs-literal">None</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter-faceid_sd15.bin&quot;</span>,
  image_encoder_folder=<span class="hljs-literal">None</span>
)
pipeline.set_ip_adapter_scale(<span class="hljs-number">0.6</span>)

image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_mask_girl1.png&quot;</span>)

ref_images_embeds = []
app = FaceAnalysis(name=<span class="hljs-string">&quot;buffalo_l&quot;</span>, providers=[<span class="hljs-string">&#x27;CUDAExecutionProvider&#x27;</span>, <span class="hljs-string">&#x27;CPUExecutionProvider&#x27;</span>])
app.prepare(ctx_id=<span class="hljs-number">0</span>, det_size=(<span class="hljs-number">640</span>, <span class="hljs-number">640</span>))
image = cv2.cvtColor(np.asarray(image), cv2.COLOR_RGB2BGR)
faces = app.get(image)
image = torch.from_numpy(faces[<span class="hljs-number">0</span>].normed_embedding)
ref_images_embeds.append(image.unsqueeze(<span class="hljs-number">0</span>))
ref_images_embeds = torch.stack(ref_images_embeds, dim=<span class="hljs-number">0</span>).unsqueeze(<span class="hljs-number">0</span>)
neg_ref_images_embeds = torch.zeros_like(ref_images_embeds)
id_embeds = torch.cat([neg_ref_images_embeds, ref_images_embeds]).to(dtype=torch.float16, device=<span class="hljs-string">&quot;cuda&quot;</span>)

pipeline(
    prompt=<span class="hljs-string">&quot;A photo of a girl&quot;</span>,
    ip_adapter_image_embeds=[id_embeds],
    negative_prompt=<span class="hljs-string">&quot;monochrome, lowres, bad anatomy, worst quality, low quality&quot;</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <p>For FaceID Plus, load <a href="https://huggingface.co/docs/transformers/main/en/model_doc/clip#transformers.CLIPVisionModelWithProjection" rel="nofollow">CLIPVisionModelWithProjection</a> as the image encoder.</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> CLIPVisionModelWithProjection
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForText2Image

image_encoder = CLIPVisionModelWithProjection.from_pretrained(
    <span class="hljs-string">&quot;laion/CLIP-ViT-H-14-laion2B-s32B-b79K&quot;</span>,
    dtype=torch.float16,
)

pipeline = AutoPipelineForText2Image.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>,
    image_encoder=image_encoder,
    dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter-FaceID&quot;</span>,
  subfolder=<span class="hljs-literal">None</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter-faceid-plus_sd15.bin&quot;</span>
)<!----></pre></div><!----> <p>FaceID Plus and Plus v2 also need CLIP image embeddings on top of the InsightFace embeds. After you prepare the face embeddings, encode the reference image and assign the CLIP embeds to the image projection layer.</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

ip_adapter_images = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_mask_girl1.png&quot;</span>)
num_images = <span class="hljs-number">1</span>
clip_embeds = pipeline.prepare_ip_adapter_image_embeds(
    ip_adapter_image=[ip_adapter_images],
    ip_adapter_image_embeds=<span class="hljs-literal">None</span>,
    device=torch.device(<span class="hljs-string">&quot;cuda&quot;</span>),
    num_images_per_prompt=num_images,
    do_classifier_free_guidance=<span class="hljs-literal">True</span>,
)[<span class="hljs-number">0</span>]

pipeline.unet.encoder_hid_proj.image_projection_layers[<span class="hljs-number">0</span>].clip_embeds = clip_embeds.to(dtype=torch.float16)
<span class="hljs-comment"># set to True if using IP-Adapter FaceID Plus v2</span>
pipeline.unet.encoder_hid_proj.image_projection_layers[<span class="hljs-number">0</span>].shortcut = <span class="hljs-literal">False</span><!----></pre></div><!----><!----></div><!----><!----></div><!----> <!--[1--><h2 class="relative group"><a id="image-embeddings" 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="#image-embeddings"><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><!---->Image embeddings<!----></span></h2><!--]--><!----> <p><code>prepare_ip_adapter_image_embeds()</code> encodes IP-Adapter images into embeddings you can save and reuse. Precompute them once instead of loading and encoding the same images every time you run the 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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForText2Image
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

pipeline = AutoPipelineForText2Image.from_pretrained(
  <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
  dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter_sdxl.bin&quot;</span>
)

image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_bear_1.png&quot;</span>)
image_embeds = pipeline.prepare_ip_adapter_image_embeds(
    ip_adapter_image=image,
    ip_adapter_image_embeds=<span class="hljs-literal">None</span>,
    device=<span class="hljs-string">&quot;cuda&quot;</span>,
    num_images_per_prompt=<span class="hljs-number">1</span>,
    do_classifier_free_guidance=<span class="hljs-literal">True</span>,
)

torch.save(image_embeds, <span class="hljs-string">&quot;image_embeds.ipadpt&quot;</span>)<!----></pre></div><!----> <p>Reload the image embeddings by passing them to the <code>ip_adapter_image_embeds</code> parameter. Set <code>image_encoder_folder</code> to <code>None</code> because you don’t need the image encoder anymore to generate the image embeddings.</p> <blockquote class="tip"><p>You can also load image embeddings from other sources such as ComfyUI.</p></blockquote> <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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->pipeline = AutoPipelineForText2Image.from_pretrained(
  <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
  dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  image_encoder_folder=<span class="hljs-literal">None</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter_sdxl.bin&quot;</span>
)
pipeline.set_ip_adapter_scale(<span class="hljs-number">0.8</span>)
image_embeds = torch.load(<span class="hljs-string">&quot;image_embeds.ipadpt&quot;</span>)
pipeline(
    prompt=<span class="hljs-string">&quot;a polar bear sitting in a chair drinking a milkshake&quot;</span>,
    ip_adapter_image_embeds=image_embeds,
    negative_prompt=<span class="hljs-string">&quot;deformed, ugly, wrong proportion, low res, bad anatomy, worst quality, low quality&quot;</span>,
    num_inference_steps=<span class="hljs-number">100</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="masking" 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="#masking"><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><!---->Masking<!----></span></h2><!--]--><!----> <p>Binary masking enables assigning an IP-Adapter image to a specific area of the output image, making it useful for composing multiple IP-Adapter images. Each IP-Adapter image requires a binary mask.</p> <p>Load the <a href="/docs/diffusers/pr_15230/en/api/loaders/ip_adapter#diffusers.IPAdapterMaskProcessor">IPAdapterMaskProcessor</a> to preprocess the image masks. For the best results, provide the output <code>height</code> and <code>width</code> to ensure masks with different aspect ratios are appropriately sized. If the input masks already match the aspect ratio of the generated image, you don’t need to set the <code>height</code> and <code>width</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForText2Image
<span class="hljs-keyword">from</span> diffusers.image_processor <span class="hljs-keyword">import</span> IPAdapterMaskProcessor
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

pipeline = AutoPipelineForText2Image.from_pretrained(
  <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
  dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

mask1 = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_mask_mask1.png&quot;</span>)
mask2 = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_mask_mask2.png&quot;</span>)

processor = IPAdapterMaskProcessor()
masks = processor.preprocess([mask1, mask2], height=<span class="hljs-number">1024</span>, width=<span class="hljs-number">1024</span>)<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/ip_mask_mask1.png" width="200" alt="mask 1"/> <figcaption style="text-align: center;">mask 1</figcaption></figure> <figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/ip_mask_mask2.png" width="200" alt="mask 2"/> <figcaption style="text-align: center;">mask 2</figcaption></figure></div> <p>Provide both the IP-Adapter images and their scales as a list. Pass the preprocessed masks to <code>cross_attention_kwargs</code> in the 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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->face_image1 = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_mask_girl1.png&quot;</span>)
face_image2 = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_mask_girl2.png&quot;</span>)

pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  weight_name=[<span class="hljs-string">&quot;ip-adapter-plus-face_sdxl_vit-h.safetensors&quot;</span>]
)
pipeline.set_ip_adapter_scale([[<span class="hljs-number">0.7</span>, <span class="hljs-number">0.7</span>]])

ip_images = [[face_image1, face_image2]]
masks = [masks.reshape(<span class="hljs-number">1</span>, masks.shape[<span class="hljs-number">0</span>], masks.shape[<span class="hljs-number">2</span>], masks.shape[<span class="hljs-number">3</span>])]

pipeline(
  prompt=<span class="hljs-string">&quot;2 girls&quot;</span>,
  ip_adapter_image=ip_images,
  negative_prompt=<span class="hljs-string">&quot;monochrome, lowres, bad anatomy, worst quality, low quality&quot;</span>,
  cross_attention_kwargs={<span class="hljs-string">&quot;ip_adapter_masks&quot;</span>: masks}
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; flex-direction: column; gap: 10px;"><div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/ip_mask_girl1.png" width="400" alt="IP-Adapter image 1"/> <figcaption style="text-align: center;">IP-Adapter image 1</figcaption></figure> <figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/ip_mask_girl2.png" width="400" alt="IP-Adapter image 2"/> <figcaption style="text-align: center;">IP-Adapter image 2</figcaption></figure></div> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_attention_mask_result_seed_0.png" width="400" alt="Generated image with mask"/> <figcaption style="text-align: center;">generated with mask</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_no_attention_mask_result_seed_0.png" width="400" alt="Generated image without mask"/> <figcaption style="text-align: center;">generated without mask</figcaption></figure></div></div> <!--[1--><h2 class="relative group"><a id="recipes" 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="#recipes"><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><!---->Recipes<!----></span></h2><!--]--><!----> <p>Combine IP-Adapter with other adapters or pipelines, like multiple adapters, ControlNet, InstantStyle, or few-step LCM, when one reference image isn’t enough.</p> <!--[2--><h3 class="relative group"><a id="multiple-ip-adapters" 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="#multiple-ip-adapters"><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><!---->Multiple IP-Adapters<!----></span></h3><!--]--><!----> <p>Combine multiple IP-Adapters to generate images in more diverse styles. For example, you can use IP-Adapter Face to generate consistent faces and characters and IP-Adapter Plus to generate those faces in specific styles.</p> <p>Load an image encoder with <a href="https://huggingface.co/docs/transformers/main/en/model_doc/clip#transformers.CLIPVisionModelWithProjection" rel="nofollow">CLIPVisionModelWithProjection</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 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="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForText2Image, DDIMScheduler
<span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> CLIPVisionModelWithProjection
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

image_encoder = CLIPVisionModelWithProjection.from_pretrained(
    <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
    subfolder=<span class="hljs-string">&quot;models/image_encoder&quot;</span>,
    dtype=torch.float16,
)<!----></pre></div><!----> <p>Load a base model, scheduler and the following IP-Adapters.</p> <ul><li><a href="https://huggingface.co/h94/IP-Adapter#ip-adapter-for-sdxl-10" rel="nofollow">ip-adapter-plus_sdxl_vit-h</a> uses patch embeddings and a ViT-H image encoder</li> <li><a href="https://huggingface.co/h94/IP-Adapter#ip-adapter-for-sdxl-10" rel="nofollow">ip-adapter-plus-face_sdxl_vit-h</a> uses patch embeddings and a ViT-H image encoder but it is conditioned on images of cropped faces</li></ul> <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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->pipeline = AutoPipelineForText2Image.from_pretrained(
    <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
    dtype=torch.float16,
    image_encoder=image_encoder,
)
pipeline.scheduler = DDIMScheduler.from_config(pipeline.scheduler.config)
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  weight_name=[<span class="hljs-string">&quot;ip-adapter-plus_sdxl_vit-h.safetensors&quot;</span>, <span class="hljs-string">&quot;ip-adapter-plus-face_sdxl_vit-h.safetensors&quot;</span>]
)
pipeline.set_ip_adapter_scale([<span class="hljs-number">0.7</span>, <span class="hljs-number">0.3</span>])
<span class="hljs-comment"># enable_model_cpu_offload to reduce memory usage</span>
pipeline.enable_model_cpu_offload()<!----></pre></div><!----> <p>Load an image and a folder containing images of a certain style to apply.</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->face_image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/women_input.png&quot;</span>)
style_folder = <span class="hljs-string">&quot;https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/style_ziggy&quot;</span>
style_images = [load_image(<span class="hljs-string">f&quot;<span class="hljs-subst">{style_folder}</span>/img<span class="hljs-subst">{i}</span>.png&quot;</span>) <span class="hljs-keyword">for</span> i <span class="hljs-keyword">in</span> <span class="hljs-built_in">range</span>(<span class="hljs-number">10</span>)]<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/women_input.png" width="400" alt="Face image"/> <figcaption style="text-align: center;">face image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/ip_style_grid.png" width="400" alt="Style images"/> <figcaption style="text-align: center;">style images</figcaption></figure></div> <p>Pass style and face images as a list to <code>ip_adapter_image</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->generator = torch.Generator(device=<span class="hljs-string">&quot;cpu&quot;</span>).manual_seed(<span class="hljs-number">0</span>)

pipeline(
    prompt=<span class="hljs-string">&quot;wonderwoman&quot;</span>,
    ip_adapter_image=[style_images, face_image],
    negative_prompt=<span class="hljs-string">&quot;monochrome, lowres, bad anatomy, worst quality, low quality&quot;</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; justify-content: center;"><figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/ip_multi_out.png" width="400" alt="Generated image"/> <figcaption style="text-align: center;">generated image</figcaption></figure></div> <!--[2--><h3 class="relative group"><a id="structural-control" 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="#structural-control"><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><!---->Structural control<!----></span></h3><!--]--><!----> <p>For structural control, combine IP-Adapter with <a href="../api/pipelines/controlnet">ControlNet</a> conditioned on depth maps, edge maps, pose estimations, and more.</p> <p>The example below loads a <a href="/docs/diffusers/pr_15230/en/api/models/controlnet#diffusers.ControlNetModel">ControlNetModel</a> checkpoint conditioned on depth maps and combines it with an IP-Adapter.</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> StableDiffusionControlNetPipeline, ControlNetModel

controlnet = ControlNetModel.from_pretrained(
  <span class="hljs-string">&quot;lllyasviel/control_v11f1p_sd15_depth&quot;</span>,
  dtype=torch.float16
)

pipeline = StableDiffusionControlNetPipeline.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>,
    controlnet=controlnet,
    dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;models&quot;</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter_sd15.bin&quot;</span>
)<!----></pre></div><!----> <p>Pass the depth map and IP-Adapter image to the 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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->depth_map = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/depth.png&quot;</span>)
ip_adapter_image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/statue.png&quot;</span>)
pipeline(
  prompt=<span class="hljs-string">&quot;best quality, high quality&quot;</span>,
  image=depth_map,
  ip_adapter_image=ip_adapter_image,
  negative_prompt=<span class="hljs-string">&quot;monochrome, lowres, bad anatomy, worst quality, low quality&quot;</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/statue.png" width="300" alt="IP-Adapter image"/> <figcaption style="text-align: center;">IP-Adapter image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/depth.png" width="300" alt="Depth map"/> <figcaption style="text-align: center;">depth map</figcaption></figure> <figure><img src="https://huggingface.co/datasets/YiYiXu/testing-images/resolve/main/ipa-controlnet-out.png" width="300" alt="Generated image"/> <figcaption style="text-align: center;">generated image</figcaption></figure></div> <!--[2--><h3 class="relative group"><a id="style-and-layout-control" 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="#style-and-layout-control"><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><!---->Style and layout control<!----></span></h3><!--]--><!----> <p>For style and layout control, combine IP-Adapter with <a href="https://huggingface.co/papers/2404.02733" rel="nofollow">InstantStyle</a>. InstantStyle separates <em>style</em> (color, texture, overall feel) from <em>content</em> and applies style only in style-specific blocks so content areas stay intact. That gives stronger style consistency and clearer layout control.</p> <p>Activate the IP-Adapter only in selected layers with <a href="/docs/diffusers/pr_15230/en/api/loaders/ip_adapter#diffusers.loaders.IPAdapterMixin.set_ip_adapter_scale">set_ip_adapter_scale()</a>. The example below turns it on in the model’s down <code>block_2</code> (layout) and up <code>block_0</code> (style).</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoPipelineForText2Image
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

pipeline = AutoPipelineForText2Image.from_pretrained(
  <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
  dtype=torch.float16
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;sdxl_models&quot;</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter_sdxl.bin&quot;</span>
)

scale = {
    <span class="hljs-string">&quot;down&quot;</span>: {<span class="hljs-string">&quot;block_2&quot;</span>: [<span class="hljs-number">0.0</span>, <span class="hljs-number">1.0</span>]},
    <span class="hljs-string">&quot;up&quot;</span>: {<span class="hljs-string">&quot;block_0&quot;</span>: [<span class="hljs-number">0.0</span>, <span class="hljs-number">1.0</span>, <span class="hljs-number">0.0</span>]},
}
pipeline.set_ip_adapter_scale(scale)<!----></pre></div><!----> <p>Load the style image and generate an image.</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->style_image = load_image(<span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/0052a70beed5bf71b92610a43a52df6d286cd5f3/diffusers/rabbit.jpg&quot;</span>)

pipeline(
    prompt=<span class="hljs-string">&quot;a cat, masterpiece, best quality, high quality&quot;</span>,
    ip_adapter_image=style_image,
    negative_prompt=<span class="hljs-string">&quot;text, watermark, lowres, low quality, worst quality, deformed, glitch, low contrast, noisy, saturation, blurry&quot;</span>,
    guidance_scale=<span class="hljs-number">5</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/0052a70beed5bf71b92610a43a52df6d286cd5f3/diffusers/rabbit.jpg" width="400" alt="Style image"/> <figcaption style="text-align: center;">style image</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/datasets/cat_style_layout.png" width="400" alt="Generated image"/> <figcaption style="text-align: center;">generated image</figcaption></figure></div> <p>The figures below compare InstantStyle’s style-only activation (up <code>block_0</code>) with turning the IP-Adapter on in all layers. All-layers usually follows the image prompt more strongly and can reduce diversity. Prefer the style-only scale when you want InstantStyle’s layout-preserving behavior.</p> <blockquote class="tip"><p>You don’t need to specify all the layers in the <code>scale</code> dictionary. Layers not included are set to 0, which means the IP-Adapter is disabled.</p></blockquote> <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> Copied</div><!----></button><!----></div> <pre class="language-py "><!---->scale = {
    <span class="hljs-string">&quot;up&quot;</span>: {<span class="hljs-string">&quot;block_0&quot;</span>: [<span class="hljs-number">0.0</span>, <span class="hljs-number">1.0</span>, <span class="hljs-number">0.0</span>]},
}
pipeline.set_ip_adapter_scale(scale)

pipeline(
    prompt=<span class="hljs-string">&quot;a cat, masterpiece, best quality, high quality&quot;</span>,
    ip_adapter_image=style_image,
    negative_prompt=<span class="hljs-string">&quot;text, watermark, lowres, low quality, worst quality, deformed, glitch, low contrast, noisy, saturation, blurry&quot;</span>,
    guidance_scale=<span class="hljs-number">5</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; gap: 10px; justify-content: space-around; align-items: flex-end;"><figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/datasets/cat_style_only.png" width="400" alt="Generated image (style only)"/> <figcaption style="text-align: center;">style-only (up block_0)</figcaption></figure> <figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/datasets/cat_ip_adapter.png" width="400" alt="Generated image (all layers)"/> <figcaption style="text-align: center;">all layers</figcaption></figure></div> <!--[2--><h3 class="relative group"><a id="instant-generation" 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="#instant-generation"><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><!---->Instant generation<!----></span></h3><!--]--><!----> <p>Combine IP-Adapter with an <a href="../api/pipelines/latent_consistency_models">LCM</a> LoRA for few-step generation.</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> Copied</div><!----></button><!----></div> <pre class="language-py "><!----><span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline, LCMScheduler
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image

pipeline = DiffusionPipeline.from_pretrained(
  <span class="hljs-string">&quot;sd-dreambooth-library/herge-style&quot;</span>,
  dtype=torch.float16
)
pipeline.load_ip_adapter(
  <span class="hljs-string">&quot;h94/IP-Adapter&quot;</span>,
  subfolder=<span class="hljs-string">&quot;models&quot;</span>,
  weight_name=<span class="hljs-string">&quot;ip-adapter_sd15.bin&quot;</span>
)
pipeline.load_lora_weights(<span class="hljs-string">&quot;latent-consistency/lcm-lora-sdv1-5&quot;</span>)
pipeline.scheduler = LCMScheduler.from_config(pipeline.scheduler.config)
pipeline.enable_model_cpu_offload()

pipeline.set_ip_adapter_scale(<span class="hljs-number">0.4</span>)
ip_adapter_image = load_image(<span class="hljs-string">&quot;https://user-images.githubusercontent.com/24734142/266492875-2d50d223-8475-44f0-a7c6-08b51cb53572.png&quot;</span>)
pipeline(
    prompt=<span class="hljs-string">&quot;herge_style woman in armor, best quality, high quality&quot;</span>,
    ip_adapter_image=ip_adapter_image,
    num_inference_steps=<span class="hljs-number">4</span>,
    guidance_scale=<span class="hljs-number">1</span>,
).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <div style="display: flex; justify-content: center;"><figure><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/ip_adapter_herge.png" width="400" alt="Generated image"/> <figcaption style="text-align: center;">generated image</figcaption></figure></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/using-diffusers/ip_adapter.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_1rjuniz = {
						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.vZc-dOUR.js"),
						import("/docs/diffusers/main/en/_app/immutable/entry/app.B6c3Ml05.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 392],
							data: [null,null],
							form: null,
							error: null
						});
					});
				}
			</script>
		
Xet Storage Details

Size:: 90.4 kB
Xet hash:: f7ebc7ea1d7419f040420cf0e03eeb5bb211ca46b4dd0b8af8f090e25db635e6
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.