diffusers/main/en/using-diffusers/freeu.html

<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Improve generation quality with FreeU&quot;,&quot;local&quot;:&quot;improve-generation-quality-with-freeu&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;StableDiffusionPipeline&quot;,&quot;local&quot;:&quot;stablediffusionpipeline&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Stable Diffusion XL&quot;,&quot;local&quot;:&quot;stable-diffusion-xl&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Text-to-video generation&quot;,&quot;local&quot;:&quot;text-to-video-generation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;: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.3eac946d.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/scheduler.182ea377.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/singletons.86eb2521.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/index.1f6d62f6.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/paths.37e3205a.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/entry/app.2ee36914.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/index.abf12888.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/nodes/0.24b761e9.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/176.0323c54e.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/Tip.230e2334.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/CodeBlock.57fe6e13.js">
		<link rel="modulepreload" href="/docs/diffusers/main/en/_app/immutable/chunks/DocNotebookDropdown.5fa27ace.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/EditOnGithub.9b8e78e4.js"><!-- HEAD_svelte-u9bgzb_START --><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Improve generation quality with FreeU&quot;,&quot;local&quot;:&quot;improve-generation-quality-with-freeu&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;StableDiffusionPipeline&quot;,&quot;local&quot;:&quot;stablediffusionpipeline&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Stable Diffusion XL&quot;,&quot;local&quot;:&quot;stable-diffusion-xl&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Text-to-video generation&quot;,&quot;local&quot;:&quot;text-to-video-generation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"><!-- HEAD_svelte-u9bgzb_END -->      <p></p>   <h1 class="relative group"><a id="improve-generation-quality-with-freeu" 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="#improve-generation-quality-with-freeu"><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>Improve generation quality with FreeU</span></h1>  <div class="flex space-x-1 absolute z-10 right-0 top-0"> <div class="relative colab-dropdown "> <button class="  " type="button">  <img alt="Open In Colab" class="!m-0" src="https://colab.research.google.com/assets/colab-badge.svg"> </button>    </div> <div class="relative colab-dropdown "> <button class="  " type="button">  <img alt="Open In Studio Lab" class="!m-0" src="https://studiolab.sagemaker.aws/studiolab.svg"> </button>    </div></div> <p data-svelte-h="svelte-1ppsh1y">The UNet is responsible for denoising during the reverse diffusion process, and there are two distinct features in its architecture:</p> <ol data-svelte-h="svelte-8wamr8"><li>Backbone features primarily contribute to the denoising process</li> <li>Skip features mainly introduce high-frequency features into the decoder module and can make the network overlook the semantics in the backbone features</li></ol> <p data-svelte-h="svelte-153lhh2">However, the skip connection can sometimes introduce unnatural image details. <a href="https://hf.co/papers/2309.11497" rel="nofollow">FreeU</a> is a technique for improving image quality by rebalancing the contributions from the UNet’s skip connections and backbone feature maps.</p> <p data-svelte-h="svelte-1l2549u">FreeU is applied during inference and it does not require any additional training. The technique works for different tasks such as text-to-image, image-to-image, and text-to-video.</p> <p data-svelte-h="svelte-1dt13h2">In this guide, you will apply FreeU to the <a href="/docs/diffusers/main/en/api/pipelines/stable_diffusion/text2img#diffusers.StableDiffusionPipeline">StableDiffusionPipeline</a>, <a href="/docs/diffusers/main/en/api/pipelines/stable_diffusion/stable_diffusion_xl#diffusers.StableDiffusionXLPipeline">StableDiffusionXLPipeline</a>, and <a href="/docs/diffusers/main/en/api/pipelines/text_to_video#diffusers.TextToVideoSDPipeline">TextToVideoSDPipeline</a>. You need to install Diffusers from source to run the examples below.</p>  <h2 class="relative group"><a id="stablediffusionpipeline" 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="#stablediffusionpipeline"><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>StableDiffusionPipeline</span></h2> <p data-svelte-h="svelte-1t6isft">Load 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 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 <span class="hljs-keyword">import</span> DiffusionPipeline
<span class="hljs-keyword">import</span> torch

pipeline = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;runwayml/stable-diffusion-v1-5&quot;</span>, torch_dtype=torch.float16, safety_checker=<span class="hljs-literal">None</span>
).to(<span class="hljs-string">&quot;cuda&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1kfovwo">Then enable the FreeU mechanism with the FreeU-specific hyperparameters. These values are scaling factors for the backbone and skip features.</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.enable_freeu(s1=<span class="hljs-number">0.9</span>, s2=<span class="hljs-number">0.2</span>, b1=<span class="hljs-number">1.2</span>, b2=<span class="hljs-number">1.4</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-gb72tw">The values above are from the official FreeU <a href="https://github.com/ChenyangSi/FreeU" rel="nofollow">code repository</a> where you can also find <a href="https://github.com/ChenyangSi/FreeU#range-for-more-parameters" rel="nofollow">reference hyperparameters</a> for different models.</p>  <div class="course-tip  bg-gradient-to-br dark:bg-gradient-to-r before:border-green-500 dark:before:border-green-800 from-green-50 dark:from-gray-900 to-white dark:to-gray-950 border border-green-50 text-green-700 dark:text-gray-400"><p data-svelte-h="svelte-vhfeba">Disable the FreeU mechanism by calling <code>disable_freeu()</code> on a pipeline.</p></div> <p data-svelte-h="svelte-yyi4rw">And then run inference:</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 -->prompt = <span class="hljs-string">&quot;A squirrel eating a burger&quot;</span>
seed = <span class="hljs-number">2023</span>
image = pipeline(prompt, generator=torch.manual_seed(seed)).images[<span class="hljs-number">0</span>]
image<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-u90fev">The figure below compares non-FreeU and FreeU results respectively for the same hyperparameters used above (<code>prompt</code> and <code>seed</code>):</p> <p data-svelte-h="svelte-gexrkr"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/freeu/sdv1_5_freeu.jpg"></p> <p data-svelte-h="svelte-61gii3">Let’s see how Stable Diffusion 2 results are impacted:</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 <span class="hljs-keyword">import</span> DiffusionPipeline
<span class="hljs-keyword">import</span> torch

pipeline = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stabilityai/stable-diffusion-2-1&quot;</span>, torch_dtype=torch.float16, safety_checker=<span class="hljs-literal">None</span>
).to(<span class="hljs-string">&quot;cuda&quot;</span>)

prompt = <span class="hljs-string">&quot;A squirrel eating a burger&quot;</span>
seed = <span class="hljs-number">2023</span>

pipeline.enable_freeu(s1=<span class="hljs-number">0.9</span>, s2=<span class="hljs-number">0.2</span>, b1=<span class="hljs-number">1.1</span>, b2=<span class="hljs-number">1.2</span>)
image = pipeline(prompt, generator=torch.manual_seed(seed)).images[<span class="hljs-number">0</span>]
image<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-evvlf0"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/freeu/sdv2_1_freeu.jpg"></p>  <h2 class="relative group"><a id="stable-diffusion-xl" 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="#stable-diffusion-xl"><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>Stable Diffusion XL</span></h2> <p data-svelte-h="svelte-2ehsm1">Finally, let’s take a look at how FreeU affects Stable Diffusion XL results:</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 <span class="hljs-keyword">import</span> DiffusionPipeline
<span class="hljs-keyword">import</span> torch

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

prompt = <span class="hljs-string">&quot;A squirrel eating a burger&quot;</span>
seed = <span class="hljs-number">2023</span>

<span class="hljs-comment"># Comes from</span>
<span class="hljs-comment"># https://wandb.ai/nasirk24/UNET-FreeU-SDXL/reports/FreeU-SDXL-Optimal-Parameters--Vmlldzo1NDg4NTUw</span>
pipeline.enable_freeu(s1=<span class="hljs-number">0.6</span>, s2=<span class="hljs-number">0.4</span>, b1=<span class="hljs-number">1.1</span>, b2=<span class="hljs-number">1.2</span>)
image = pipeline(prompt, generator=torch.manual_seed(seed)).images[<span class="hljs-number">0</span>]
image<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1vig68s"><img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/freeu/sdxl_freeu.jpg"></p>  <h2 class="relative group"><a id="text-to-video-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="#text-to-video-generation"><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>Text-to-video generation</span></h2> <p data-svelte-h="svelte-pga2sp">FreeU can also be used to improve video quality:</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 <span class="hljs-keyword">import</span> DiffusionPipeline
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> export_to_video
<span class="hljs-keyword">import</span> torch

model_id = <span class="hljs-string">&quot;cerspense/zeroscope_v2_576w&quot;</span>
pipe = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16).to(<span class="hljs-string">&quot;cuda&quot;</span>)

prompt = <span class="hljs-string">&quot;an astronaut riding a horse on mars&quot;</span>
seed = <span class="hljs-number">2023</span>

<span class="hljs-comment"># The values come from</span>
<span class="hljs-comment"># https://github.com/lyn-rgb/FreeU_Diffusers#video-pipelines</span>
pipe.enable_freeu(b1=<span class="hljs-number">1.2</span>, b2=<span class="hljs-number">1.4</span>, s1=<span class="hljs-number">0.9</span>, s2=<span class="hljs-number">0.2</span>)
video_frames = pipe(prompt, height=<span class="hljs-number">320</span>, width=<span class="hljs-number">576</span>, num_frames=<span class="hljs-number">30</span>, generator=torch.manual_seed(seed)).frames[<span class="hljs-number">0</span>]
export_to_video(video_frames, <span class="hljs-string">&quot;astronaut_rides_horse.mp4&quot;</span>)<!-- HTML_TAG_END --></pre></div> <p data-svelte-h="svelte-1cx6l39">Thanks to <a href="https://github.com/kadirnar/" rel="nofollow">kadirnar</a> for helping to integrate the feature, and to <a href="https://github.com/justindujardin" rel="nofollow">justindujardin</a> for the helpful discussions.</p> <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/freeu.md" target="_blank"><span data-svelte-h="svelte-1kd6by1">&lt;</span> <span data-svelte-h="svelte-x0xyl0">&gt;</span> <span data-svelte-h="svelte-1dajgef"><span class="underline ml-1.5">Update</span> on GitHub</span></a>  <p></p> 
			
			<script>
				{
					__sveltekit_5j5g7y = {
						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.3eac946d.js"),
						import("/docs/diffusers/main/en/_app/immutable/entry/app.2ee36914.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 176],
							data,
							form: null,
							error: null
						});
					});
				}
			</script>