diffusers/pr_11986/en/_app/immutable/nodes/305.55ff0614.js

import{s as Zs,o as Js,n as he}from"../chunks/scheduler.8c3d61f6.js";import{S as Ts,i as $s,g as f,s as i,r as h,A as js,h as m,f as l,c as r,j as Us,u as M,x as U,k as Oe,y as vs,a as t,v as y,d as b,t as w,w as g}from"../chunks/index.da70eac4.js";import{T as Gs}from"../chunks/Tip.1d9b8c37.js";import{C as T}from"../chunks/CodeBlock.a9c4becf.js";import{D as _s}from"../chunks/DocNotebookDropdown.48852948.js";import{H as me,E as Ws}from"../chunks/getInferenceSnippets.366c2c95.js";import{H as Vs,a as es}from"../chunks/HfOption.6ab18950.js";function ks($){let n,Z='<a href="/docs/diffusers/pr_11986/en/api/schedulers/lms_discrete#diffusers.LMSDiscreteScheduler">LMSDiscreteScheduler</a> typically generates higher quality images than the default scheduler.',u,o,c;return o=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMExNU0Rpc2NyZXRlU2NoZWR1bGVyJTBBJTBBcGlwZWxpbmUuc2NoZWR1bGVyJTIwJTNEJTIwTE1TRGlzY3JldGVTY2hlZHVsZXIuZnJvbV9jb25maWcocGlwZWxpbmUuc2NoZWR1bGVyLmNvbmZpZyklMEFpbWFnZSUyMCUzRCUyMHBpcGVsaW5lKHByb21wdCUyQyUyMGdlbmVyYXRvciUzRGdlbmVyYXRvcikuaW1hZ2VzJTVCMCU1RCUwQWltYWdl",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> LMSDiscreteScheduler

pipeline.scheduler = LMSDiscreteScheduler.from_config(pipeline.scheduler.config)
image = pipeline(prompt, generator=generator).images[<span class="hljs-number">0</span>]
image`,wrap:!1}}),{c(){n=f("p"),n.innerHTML=Z,u=i(),h(o.$$.fragment)},l(a){n=m(a,"P",{"data-svelte-h":!0}),U(n)!=="svelte-12ncfeg"&&(n.innerHTML=Z),u=r(a),M(o.$$.fragment,a)},m(a,d){t(a,n,d),t(a,u,d),y(o,a,d),c=!0},p:he,i(a){c||(b(o.$$.fragment,a),c=!0)},o(a){w(o.$$.fragment,a),c=!1},d(a){a&&(l(n),l(u)),g(o,a)}}}function Ss($){let n,Z='<a href="/docs/diffusers/pr_11986/en/api/schedulers/euler#diffusers.EulerDiscreteScheduler">EulerDiscreteScheduler</a> can generate higher quality images in just 30 steps.',u,o,c;return o=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMEV1bGVyRGlzY3JldGVTY2hlZHVsZXIlMEElMEFwaXBlbGluZS5zY2hlZHVsZXIlMjAlM0QlMjBFdWxlckRpc2NyZXRlU2NoZWR1bGVyLmZyb21fY29uZmlnKHBpcGVsaW5lLnNjaGVkdWxlci5jb25maWcpJTBBaW1hZ2UlMjAlM0QlMjBwaXBlbGluZShwcm9tcHQlMkMlMjBnZW5lcmF0b3IlM0RnZW5lcmF0b3IpLmltYWdlcyU1QjAlNUQlMEFpbWFnZQ==",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> EulerDiscreteScheduler

pipeline.scheduler = EulerDiscreteScheduler.from_config(pipeline.scheduler.config)
image = pipeline(prompt, generator=generator).images[<span class="hljs-number">0</span>]
image`,wrap:!1}}),{c(){n=f("p"),n.innerHTML=Z,u=i(),h(o.$$.fragment)},l(a){n=m(a,"P",{"data-svelte-h":!0}),U(n)!=="svelte-nf0soq"&&(n.innerHTML=Z),u=r(a),M(o.$$.fragment,a)},m(a,d){t(a,n,d),t(a,u,d),y(o,a,d),c=!0},p:he,i(a){c||(b(o.$$.fragment,a),c=!0)},o(a){w(o.$$.fragment,a),c=!1},d(a){a&&(l(n),l(u)),g(o,a)}}}function xs($){let n,Z='<a href="/docs/diffusers/pr_11986/en/api/schedulers/euler_ancestral#diffusers.EulerAncestralDiscreteScheduler">EulerAncestralDiscreteScheduler</a> can generate higher quality images in just 30 steps.',u,o,c;return o=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMEV1bGVyQW5jZXN0cmFsRGlzY3JldGVTY2hlZHVsZXIlMEElMEFwaXBlbGluZS5zY2hlZHVsZXIlMjAlM0QlMjBFdWxlckFuY2VzdHJhbERpc2NyZXRlU2NoZWR1bGVyLmZyb21fY29uZmlnKHBpcGVsaW5lLnNjaGVkdWxlci5jb25maWcpJTBBaW1hZ2UlMjAlM0QlMjBwaXBlbGluZShwcm9tcHQlMkMlMjBnZW5lcmF0b3IlM0RnZW5lcmF0b3IpLmltYWdlcyU1QjAlNUQlMEFpbWFnZQ==",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> EulerAncestralDiscreteScheduler

pipeline.scheduler = EulerAncestralDiscreteScheduler.from_config(pipeline.scheduler.config)
image = pipeline(prompt, generator=generator).images[<span class="hljs-number">0</span>]
image`,wrap:!1}}),{c(){n=f("p"),n.innerHTML=Z,u=i(),h(o.$$.fragment)},l(a){n=m(a,"P",{"data-svelte-h":!0}),U(n)!=="svelte-1ui6p9o"&&(n.innerHTML=Z),u=r(a),M(o.$$.fragment,a)},m(a,d){t(a,n,d),t(a,u,d),y(o,a,d),c=!0},p:he,i(a){c||(b(o.$$.fragment,a),c=!0)},o(a){w(o.$$.fragment,a),c=!1},d(a){a&&(l(n),l(u)),g(o,a)}}}function Cs($){let n,Z='<a href="/docs/diffusers/pr_11986/en/api/schedulers/multistep_dpm_solver#diffusers.DPMSolverMultistepScheduler">DPMSolverMultistepScheduler</a> provides a balance between speed and quality and can generate higher quality images in just 20 steps.',u,o,c;return o=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMERQTVNvbHZlck11bHRpc3RlcFNjaGVkdWxlciUwQSUwQXBpcGVsaW5lLnNjaGVkdWxlciUyMCUzRCUyMERQTVNvbHZlck11bHRpc3RlcFNjaGVkdWxlci5mcm9tX2NvbmZpZyhwaXBlbGluZS5zY2hlZHVsZXIuY29uZmlnKSUwQWltYWdlJTIwJTNEJTIwcGlwZWxpbmUocHJvbXB0JTJDJTIwZ2VuZXJhdG9yJTNEZ2VuZXJhdG9yKS5pbWFnZXMlNUIwJTVEJTBBaW1hZ2U=",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DPMSolverMultistepScheduler

pipeline.scheduler = DPMSolverMultistepScheduler.from_config(pipeline.scheduler.config)
image = pipeline(prompt, generator=generator).images[<span class="hljs-number">0</span>]
image`,wrap:!1}}),{c(){n=f("p"),n.innerHTML=Z,u=i(),h(o.$$.fragment)},l(a){n=m(a,"P",{"data-svelte-h":!0}),U(n)!=="svelte-r5kg7i"&&(n.innerHTML=Z),u=r(a),M(o.$$.fragment,a)},m(a,d){t(a,n,d),t(a,u,d),y(o,a,d),c=!0},p:he,i(a){c||(b(o.$$.fragment,a),c=!0)},o(a){w(o.$$.fragment,a),c=!1},d(a){a&&(l(n),l(u)),g(o,a)}}}function Is($){let n,Z,u,o,c,a,d,v;return n=new es({props:{id:"schedulers",option:"LMSDiscreteScheduler",$$slots:{default:[ks]},$$scope:{ctx:$}}}),u=new es({props:{id:"schedulers",option:"EulerDiscreteScheduler",$$slots:{default:[Ss]},$$scope:{ctx:$}}}),c=new es({props:{id:"schedulers",option:"EulerAncestralDiscreteScheduler",$$slots:{default:[xs]},$$scope:{ctx:$}}}),d=new es({props:{id:"schedulers",option:"DPMSolverMultistepScheduler",$$slots:{default:[Cs]},$$scope:{ctx:$}}}),{c(){h(n.$$.fragment),Z=i(),h(u.$$.fragment),o=i(),h(c.$$.fragment),a=i(),h(d.$$.fragment)},l(p){M(n.$$.fragment,p),Z=r(p),M(u.$$.fragment,p),o=r(p),M(c.$$.fragment,p),a=r(p),M(d.$$.fragment,p)},m(p,J){y(n,p,J),t(p,Z,J),y(u,p,J),t(p,o,J),y(c,p,J),t(p,a,J),y(d,p,J),v=!0},p(p,J){const k={};J&2&&(k.$$scope={dirty:J,ctx:p}),n.$set(k);const j={};J&2&&(j.$$scope={dirty:J,ctx:p}),u.$set(j);const de={};J&2&&(de.$$scope={dirty:J,ctx:p}),c.$set(de);const S={};J&2&&(S.$$scope={dirty:J,ctx:p}),d.$set(S)},i(p){v||(b(n.$$.fragment,p),b(u.$$.fragment,p),b(c.$$.fragment,p),b(d.$$.fragment,p),v=!0)},o(p){w(n.$$.fragment,p),w(u.$$.fragment,p),w(c.$$.fragment,p),w(d.$$.fragment,p),v=!1},d(p){p&&(l(Z),l(o),l(a)),g(n,p),g(u,p),g(c,p),g(d,p)}}}function Rs($){let n,Z='The <code>FlaxLMSDiscreteScheduler</code> and <code>FlaxDDPMScheduler</code> are not compatible with the <a href="/docs/diffusers/pr_11986/en/api/pipelines/stable_diffusion/text2img#diffusers.FlaxStableDiffusionPipeline">FlaxStableDiffusionPipeline</a> yet.';return{c(){n=f("p"),n.innerHTML=Z},l(u){n=m(u,"P",{"data-svelte-h":!0}),U(n)!=="svelte-1jaa7kg"&&(n.innerHTML=Z)},m(u,o){t(u,n,o)},p:he,d(u){u&&l(n)}}}function Bs($){let n,Z,u,o,c,a,d,v,p,J="Diffusion pipelines are a collection of interchangeable schedulers and models that can be mixed and matched to tailor a pipeline to a specific use case. The scheduler encapsulates the entire denoising process such as the number of denoising steps and the algorithm for finding the denoised sample. A scheduler is not parameterized or trained so they don’t take very much memory. The model is usually only concerned with the forward pass of going from a noisy input to a less noisy sample.",k,j,de='This guide will show you how to load schedulers and models to customize a pipeline. You’ll use the <a href="https://hf.co/stable-diffusion-v1-5/stable-diffusion-v1-5" rel="nofollow">stable-diffusion-v1-5/stable-diffusion-v1-5</a> checkpoint throughout this guide, so let’s load it first.',S,x,Me,C,ss="You can see what scheduler this pipeline uses with the <code>pipeline.scheduler</code> attribute.",ye,I,be,R,we,B,ls='Schedulers are defined by a configuration file that can be used by a variety of schedulers. Load a scheduler with the <a href="/docs/diffusers/pr_11986/en/api/schedulers/overview#diffusers.SchedulerMixin.from_pretrained">SchedulerMixin.from_pretrained()</a> method, and specify the <code>subfolder</code> parameter to load the configuration file into the correct subfolder of the pipeline repository.',ge,X,ts='For example, to load the <a href="/docs/diffusers/pr_11986/en/api/schedulers/ddim#diffusers.DDIMScheduler">DDIMScheduler</a>:',Ue,E,Ze,F,as="Then you can pass the newly loaded scheduler to the pipeline.",Je,H,Te,N,$e,L,ns="Schedulers have their own unique strengths and weaknesses, making it difficult to quantitatively compare which scheduler works best for a pipeline. You typically have to make a trade-off between denoising speed and denoising quality. We recommend trying out different schedulers to find one that works best for your use case. Call the <code>pipeline.scheduler.compatibles</code> attribute to see what schedulers are compatible with a pipeline.",je,Q,is='Let’s compare the <a href="/docs/diffusers/pr_11986/en/api/schedulers/lms_discrete#diffusers.LMSDiscreteScheduler">LMSDiscreteScheduler</a>, <a href="/docs/diffusers/pr_11986/en/api/schedulers/euler#diffusers.EulerDiscreteScheduler">EulerDiscreteScheduler</a>, <a href="/docs/diffusers/pr_11986/en/api/schedulers/euler_ancestral#diffusers.EulerAncestralDiscreteScheduler">EulerAncestralDiscreteScheduler</a>, and the <a href="/docs/diffusers/pr_11986/en/api/schedulers/multistep_dpm_solver#diffusers.DPMSolverMultistepScheduler">DPMSolverMultistepScheduler</a> on the following prompt and seed.',ve,z,Ge,q,rs='To change the pipelines scheduler, use the <a href="/docs/diffusers/pr_11986/en/api/configuration#diffusers.ConfigMixin.from_config">from_config()</a> method to load a different scheduler’s <code>pipeline.scheduler.config</code> into the pipeline.',_e,G,We,_,ps='<div><img class="rounded-xl" src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_lms.png"/> <figcaption class="mt-2 text-center text-sm text-gray-500">LMSDiscreteScheduler</figcaption></div> <div><img class="rounded-xl" src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_euler_discrete.png"/> <figcaption class="mt-2 text-center text-sm text-gray-500">EulerDiscreteScheduler</figcaption></div>',Ve,W,os='<div><img class="rounded-xl" src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_euler_ancestral.png"/> <figcaption class="mt-2 text-center text-sm text-gray-500">EulerAncestralDiscreteScheduler</figcaption></div> <div><img class="rounded-xl" src="https://huggingface.co/datasets/patrickvonplaten/images/resolve/main/diffusers_docs/astronaut_dpm.png"/> <figcaption class="mt-2 text-center text-sm text-gray-500">DPMSolverMultistepScheduler</figcaption></div>',ke,D,us="Most images look very similar and are comparable in quality. Again, it often comes down to your specific use case so a good approach is to run multiple different schedulers and compare the results.",Se,Y,xe,P,cs='To compare Flax schedulers, you need to additionally load the scheduler state into the model parameters. For example, let’s change the default scheduler in <a href="/docs/diffusers/pr_11986/en/api/pipelines/stable_diffusion/text2img#diffusers.FlaxStableDiffusionPipeline">FlaxStableDiffusionPipeline</a> to use the super fast <code>FlaxDPMSolverMultistepScheduler</code>.',Ce,V,Ie,A,Re,K,ds="Then you can take advantage of Flax’s compatibility with TPUs to generate a number of images in parallel. You’ll need to make a copy of the model parameters for each available device and then split the inputs across them to generate your desired number of images.",Be,O,Xe,ee,Ee,se,fs='Models are loaded from the <a href="/docs/diffusers/pr_11986/en/api/models/overview#diffusers.ModelMixin.from_pretrained">ModelMixin.from_pretrained()</a> method, which downloads and caches the latest version of the model weights and configurations. If the latest files are available in the local cache, <a href="/docs/diffusers/pr_11986/en/api/models/overview#diffusers.ModelMixin.from_pretrained">from_pretrained()</a> reuses files in the cache instead of re-downloading them.',Fe,le,ms='Models can be loaded from a subfolder with the <code>subfolder</code> argument. For example, the model weights for <a href="https://hf.co/stable-diffusion-v1-5/stable-diffusion-v1-5" rel="nofollow">stable-diffusion-v1-5/stable-diffusion-v1-5</a> are stored in the <a href="https://hf.co/stable-diffusion-v1-5/stable-diffusion-v1-5/tree/main/unet" rel="nofollow">unet</a> subfolder.',He,te,Ne,ae,hs='They can also be directly loaded from a <a href="https://huggingface.co/google/ddpm-cifar10-32/tree/main" rel="nofollow">repository</a>.',Le,ne,Qe,ie,Ms='To load and save model variants, specify the <code>variant</code> argument in <a href="/docs/diffusers/pr_11986/en/api/models/overview#diffusers.ModelMixin.from_pretrained">ModelMixin.from_pretrained()</a> and <a href="/docs/diffusers/pr_11986/en/api/models/overview#diffusers.ModelMixin.save_pretrained">ModelMixin.save_pretrained()</a>.',ze,re,qe,pe,ys='Use the <code>torch_dtype</code> argument in <a href="/docs/diffusers/pr_11986/en/api/models/overview#diffusers.ModelMixin.from_pretrained">from_pretrained()</a> to specify the dtype to load a model in.',De,oe,Ye,ue,bs='You can also use the <a href="https://docs.pytorch.org/docs/stable/generated/torch.Tensor.to.html" rel="nofollow">torch.Tensor.to</a> method to convert to the specified dtype on the fly. It converts <em>all</em> weights unlike the <code>torch_dtype</code> argument that respects the <code>_keep_in_fp32_modules</code>. This is important for models whose layers must remain in fp32 for numerical stability and best generation quality (see example <a href="https://github.com/huggingface/diffusers/blob/f864a9a352fa4a220d860bfdd1782e3e5af96382/src/diffusers/models/transformers/transformer_wan.py#L374" rel="nofollow">here</a>).',Pe,ce,Ae,fe,Ke;return c=new me({props:{title:"Load schedulers and models",local:"load-schedulers-and-models",headingTag:"h1"}}),d=new _s({props:{classNames:"absolute z-10 right-0 top-0",options:[{label:"Mixed",value:"https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers_doc/en/schedulers.ipynb"},{label:"PyTorch",value:"https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers_doc/en/pytorch/schedulers.ipynb"},{label:"TensorFlow",value:"https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers_doc/en/tensorflow/schedulers.ipynb"},{label:"Mixed",value:"https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/main/diffusers_doc/en/schedulers.ipynb"},{label:"PyTorch",value:"https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/main/diffusers_doc/en/pytorch/schedulers.ipynb"},{label:"TensorFlow",value:"https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/main/diffusers_doc/en/tensorflow/schedulers.ipynb"}]}}),x=new T({props:{code:"aW1wb3J0JTIwdG9yY2glMEFmcm9tJTIwZGlmZnVzZXJzJTIwaW1wb3J0JTIwRGlmZnVzaW9uUGlwZWxpbmUlMEElMEFwaXBlbGluZSUyMCUzRCUyMERpZmZ1c2lvblBpcGVsaW5lLmZyb21fcHJldHJhaW5lZCglMEElMjAlMjAlMjAlMjAlMjJzdGFibGUtZGlmZnVzaW9uLXYxLTUlMkZzdGFibGUtZGlmZnVzaW9uLXYxLTUlMjIlMkMlMjB0b3JjaF9kdHlwZSUzRHRvcmNoLmZsb2F0MTYlMkMlMjB1c2Vfc2FmZXRlbnNvcnMlM0RUcnVlJTBBKS50byglMjJjdWRhJTIyKQ==",highlighted:`<span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

pipeline = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>, torch_dtype=torch.float16, use_safetensors=<span class="hljs-literal">True</span>
).to(<span class="hljs-string">&quot;cuda&quot;</span>)`,wrap:!1}}),I=new T({props:{code:"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",highlighted:`pipeline.scheduler
PNDMScheduler {
  <span class="hljs-string">&quot;_class_name&quot;</span>: <span class="hljs-string">&quot;PNDMScheduler&quot;</span>,
  <span class="hljs-string">&quot;_diffusers_version&quot;</span>: <span class="hljs-string">&quot;0.21.4&quot;</span>,
  <span class="hljs-string">&quot;beta_end&quot;</span>: <span class="hljs-number">0.012</span>,
  <span class="hljs-string">&quot;beta_schedule&quot;</span>: <span class="hljs-string">&quot;scaled_linear&quot;</span>,
  <span class="hljs-string">&quot;beta_start&quot;</span>: <span class="hljs-number">0.00085</span>,
  <span class="hljs-string">&quot;clip_sample&quot;</span>: false,
  <span class="hljs-string">&quot;num_train_timesteps&quot;</span>: <span class="hljs-number">1000</span>,
  <span class="hljs-string">&quot;set_alpha_to_one&quot;</span>: false,
  <span class="hljs-string">&quot;skip_prk_steps&quot;</span>: true,
  <span class="hljs-string">&quot;steps_offset&quot;</span>: <span class="hljs-number">1</span>,
  <span class="hljs-string">&quot;timestep_spacing&quot;</span>: <span class="hljs-string">&quot;leading&quot;</span>,
  <span class="hljs-string">&quot;trained_betas&quot;</span>: null
}`,wrap:!1}}),R=new me({props:{title:"Load a scheduler",local:"load-a-scheduler",headingTag:"h2"}}),E=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMERESU1TY2hlZHVsZXIlMkMlMjBEaWZmdXNpb25QaXBlbGluZSUwQSUwQWRkaW0lMjAlM0QlMjBERElNU2NoZWR1bGVyLmZyb21fcHJldHJhaW5lZCglMjJzdGFibGUtZGlmZnVzaW9uLXYxLTUlMkZzdGFibGUtZGlmZnVzaW9uLXYxLTUlMjIlMkMlMjBzdWJmb2xkZXIlM0QlMjJzY2hlZHVsZXIlMjIp",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DDIMScheduler, DiffusionPipeline

ddim = DDIMScheduler.from_pretrained(<span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>, subfolder=<span class="hljs-string">&quot;scheduler&quot;</span>)`,wrap:!1}}),H=new T({props:{code:"cGlwZWxpbmUlMjAlM0QlMjBEaWZmdXNpb25QaXBlbGluZS5mcm9tX3ByZXRyYWluZWQoJTBBJTIwJTIwJTIwJTIwJTIyc3RhYmxlLWRpZmZ1c2lvbi12MS01JTJGc3RhYmxlLWRpZmZ1c2lvbi12MS01JTIyJTJDJTIwc2NoZWR1bGVyJTNEZGRpbSUyQyUyMHRvcmNoX2R0eXBlJTNEdG9yY2guZmxvYXQxNiUyQyUyMHVzZV9zYWZldGVuc29ycyUzRFRydWUlMEEpLnRvKCUyMmN1ZGElMjIp",highlighted:`pipeline = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>, scheduler=ddim, torch_dtype=torch.float16, use_safetensors=<span class="hljs-literal">True</span>
).to(<span class="hljs-string">&quot;cuda&quot;</span>)`,wrap:!1}}),N=new me({props:{title:"Compare schedulers",local:"compare-schedulers",headingTag:"h2"}}),z=new T({props:{code:"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",highlighted:`<span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

pipeline = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>, torch_dtype=torch.float16, use_safetensors=<span class="hljs-literal">True</span>
).to(<span class="hljs-string">&quot;cuda&quot;</span>)

prompt = <span class="hljs-string">&quot;A photograph of an astronaut riding a horse on Mars, high resolution, high definition.&quot;</span>
generator = torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">8</span>)`,wrap:!1}}),G=new Vs({props:{id:"schedulers",options:["LMSDiscreteScheduler","EulerDiscreteScheduler","EulerAncestralDiscreteScheduler","DPMSolverMultistepScheduler"],$$slots:{default:[Is]},$$scope:{ctx:$}}}),Y=new me({props:{title:"Flax schedulers",local:"flax-schedulers",headingTag:"h3"}}),V=new Gs({props:{warning:!0,$$slots:{default:[Rs]},$$scope:{ctx:$}}}),A=new T({props:{code:"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",highlighted:`<span class="hljs-keyword">import</span> jax
<span class="hljs-keyword">import</span> numpy <span class="hljs-keyword">as</span> np
<span class="hljs-keyword">from</span> flax.jax_utils <span class="hljs-keyword">import</span> replicate
<span class="hljs-keyword">from</span> flax.training.common_utils <span class="hljs-keyword">import</span> shard
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> FlaxStableDiffusionPipeline, FlaxDPMSolverMultistepScheduler

scheduler, scheduler_state = FlaxDPMSolverMultistepScheduler.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>,
    subfolder=<span class="hljs-string">&quot;scheduler&quot;</span>
)
pipeline, params = FlaxStableDiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>,
    scheduler=scheduler,
    variant=<span class="hljs-string">&quot;bf16&quot;</span>,
    dtype=jax.numpy.bfloat16,
)
params[<span class="hljs-string">&quot;scheduler&quot;</span>] = scheduler_state`,wrap:!1}}),O=new T({props:{code:"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",highlighted:`<span class="hljs-comment"># Generate 1 image per parallel device (8 on TPUv2-8 or TPUv3-8)</span>
prompt = <span class="hljs-string">&quot;A photograph of an astronaut riding a horse on Mars, high resolution, high definition.&quot;</span>
num_samples = jax.device_count()
prompt_ids = pipeline.prepare_inputs([prompt] * num_samples)

prng_seed = jax.random.PRNGKey(<span class="hljs-number">0</span>)
num_inference_steps = <span class="hljs-number">25</span>

<span class="hljs-comment"># shard inputs and rng</span>
params = replicate(params)
prng_seed = jax.random.split(prng_seed, jax.device_count())
prompt_ids = shard(prompt_ids)

images = pipeline(prompt_ids, params, prng_seed, num_inference_steps, jit=<span class="hljs-literal">True</span>).images
images = pipeline.numpy_to_pil(np.asarray(images.reshape((num_samples,) + images.shape[-<span class="hljs-number">3</span>:])))`,wrap:!1}}),ee=new me({props:{title:"Models",local:"models",headingTag:"h2"}}),te=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMFVOZXQyRENvbmRpdGlvbk1vZGVsJTBBJTBBdW5ldCUyMCUzRCUyMFVOZXQyRENvbmRpdGlvbk1vZGVsLmZyb21fcHJldHJhaW5lZCglMjJzdGFibGUtZGlmZnVzaW9uLXYxLTUlMkZzdGFibGUtZGlmZnVzaW9uLXYxLTUlMjIlMkMlMjBzdWJmb2xkZXIlM0QlMjJ1bmV0JTIyJTJDJTIwdXNlX3NhZmV0ZW5zb3JzJTNEVHJ1ZSk=",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> UNet2DConditionModel

unet = UNet2DConditionModel.from_pretrained(<span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>, subfolder=<span class="hljs-string">&quot;unet&quot;</span>, use_safetensors=<span class="hljs-literal">True</span>)`,wrap:!1}}),ne=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMFVOZXQyRE1vZGVsJTBBJTBBdW5ldCUyMCUzRCUyMFVOZXQyRE1vZGVsLmZyb21fcHJldHJhaW5lZCglMjJnb29nbGUlMkZkZHBtLWNpZmFyMTAtMzIlMjIlMkMlMjB1c2Vfc2FmZXRlbnNvcnMlM0RUcnVlKQ==",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> UNet2DModel

unet = UNet2DModel.from_pretrained(<span class="hljs-string">&quot;google/ddpm-cifar10-32&quot;</span>, use_safetensors=<span class="hljs-literal">True</span>)`,wrap:!1}}),re=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMFVOZXQyRENvbmRpdGlvbk1vZGVsJTBBJTBBdW5ldCUyMCUzRCUyMFVOZXQyRENvbmRpdGlvbk1vZGVsLmZyb21fcHJldHJhaW5lZCglMEElMjAlMjAlMjAlMjAlMjJzdGFibGUtZGlmZnVzaW9uLXYxLTUlMkZzdGFibGUtZGlmZnVzaW9uLXYxLTUlMjIlMkMlMjBzdWJmb2xkZXIlM0QlMjJ1bmV0JTIyJTJDJTIwdmFyaWFudCUzRCUyMm5vbl9lbWElMjIlMkMlMjB1c2Vfc2FmZXRlbnNvcnMlM0RUcnVlJTBBKSUwQXVuZXQuc2F2ZV9wcmV0cmFpbmVkKCUyMi4lMkZsb2NhbC11bmV0JTIyJTJDJTIwdmFyaWFudCUzRCUyMm5vbl9lbWElMjIp",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> UNet2DConditionModel

unet = UNet2DConditionModel.from_pretrained(
    <span class="hljs-string">&quot;stable-diffusion-v1-5/stable-diffusion-v1-5&quot;</span>, subfolder=<span class="hljs-string">&quot;unet&quot;</span>, variant=<span class="hljs-string">&quot;non_ema&quot;</span>, use_safetensors=<span class="hljs-literal">True</span>
)
unet.save_pretrained(<span class="hljs-string">&quot;./local-unet&quot;</span>, variant=<span class="hljs-string">&quot;non_ema&quot;</span>)`,wrap:!1}}),oe=new T({props:{code:"ZnJvbSUyMGRpZmZ1c2VycyUyMGltcG9ydCUyMEF1dG9Nb2RlbCUwQSUwQXVuZXQlMjAlM0QlMjBBdXRvTW9kZWwuZnJvbV9wcmV0cmFpbmVkKCUwQSUyMCUyMCUyMCUyMCUyMnN0YWJpbGl0eWFpJTJGc3RhYmxlLWRpZmZ1c2lvbi14bC1iYXNlLTEuMCUyMiUyQyUyMHN1YmZvbGRlciUzRCUyMnVuZXQlMjIlMkMlMjB0b3JjaF9kdHlwZSUzRHRvcmNoLmZsb2F0MTYlMEEp",highlighted:`<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> AutoModel

unet = AutoModel.from_pretrained(
    <span class="hljs-string">&quot;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>, subfolder=<span class="hljs-string">&quot;unet&quot;</span>, torch_dtype=torch.float16
)`,wrap:!1}}),ce=new Ws({props:{source:"https://github.com/huggingface/diffusers/blob/main/docs/source/en/using-diffusers/schedulers.md"}}),{c(){n=f("meta"),Z=i(),u=f("p"),o=i(),h(c.$$.fragment),a=i(),h(d.$$.fragment),v=i(),p=f("p"),p.textContent=J,k=i(),j=f("p"),j.innerHTML=de,S=i(),h(x.$$.fragment),Me=i(),C=f("p"),C.innerHTML=ss,ye=i(),h(I.$$.fragment),be=i(),h(R.$$.fragment),we=i(),B=f("p"),B.innerHTML=ls,ge=i(),X=f("p"),X.innerHTML=ts,Ue=i(),h(E.$$.fragment),Ze=i(),F=f("p"),F.textContent=as,Je=i(),h(H.$$.fragment),Te=i(),h(N.$$.fragment),$e=i(),L=f("p"),L.innerHTML=ns,je=i(),Q=f("p"),Q.innerHTML=is,ve=i(),h(z.$$.fragment),Ge=i(),q=f("p"),q.innerHTML=rs,_e=i(),h(G.$$.fragment),We=i(),_=f("div"),_.innerHTML=ps,Ve=i(),W=f("div"),W.innerHTML=os,ke=i(),D=f("p"),D.textContent=us,Se=i(),h(Y.$$.fragment),xe=i(),P=f("p"),P.innerHTML=cs,Ce=i(),h(V.$$.fragment),Ie=i(),h(A.$$.fragment),Re=i(),K=f("p"),K.textContent=ds,Be=i(),h(O.$$.fragment),Xe=i(),h(ee.$$.fragment),Ee=i(),se=f("p"),se.innerHTML=fs,Fe=i(),le=f("p"),le.innerHTML=ms,He=i(),h(te.$$.fragment),Ne=i(),ae=f("p"),ae.innerHTML=hs,Le=i(),h(ne.$$.fragment),Qe=i(),ie=f("p"),ie.innerHTML=Ms,ze=i(),h(re.$$.fragment),qe=i(),pe=f("p"),pe.innerHTML=ys,De=i(),h(oe.$$.fragment),Ye=i(),ue=f("p"),ue.innerHTML=bs,Pe=i(),h(ce.$$.fragment),Ae=i(),fe=f("p"),this.h()},l(e){const s=js("svelte-u9bgzb",document.head);n=m(s,"META",{name:!0,content:!0}),s.forEach(l),Z=r(e),u=m(e,"P",{}),Us(u).forEach(l),o=r(e),M(c.$$.fragment,e),a=r(e),M(d.$$.fragment,e),v=r(e),p=m(e,"P",{"data-svelte-h":!0}),U(p)!=="svelte-147f8qd"&&(p.textContent=J),k=r(e),j=m(e,"P",{"data-svelte-h":!0}),U(j)!=="svelte-1vfyhb5"&&(j.innerHTML=de),S=r(e),M(x.$$.fragment,e),Me=r(e),C=m(e,"P",{"data-svelte-h":!0}),U(C)!=="svelte-bntjs"&&(C.innerHTML=ss),ye=r(e),M(I.$$.fragment,e),be=r(e),M(R.$$.fragment,e),we=r(e),B=m(e,"P",{"data-svelte-h":!0}),U(B)!=="svelte-1phbiea"&&(B.innerHTML=ls),ge=r(e),X=m(e,"P",{"data-svelte-h":!0}),U(X)!=="svelte-1j61ony"&&(X.innerHTML=ts),Ue=r(e),M(E.$$.fragment,e),Ze=r(e),F=m(e,"P",{"data-svelte-h":!0}),U(F)!=="svelte-1gk96nj"&&(F.textContent=as),Je=r(e),M(H.$$.fragment,e),Te=r(e),M(N.$$.fragment,e),$e=r(e),L=m(e,"P",{"data-svelte-h":!0}),U(L)!=="svelte-gbqn3x"&&(L.innerHTML=ns),je=r(e),Q=m(e,"P",{"data-svelte-h":!0}),U(Q)!=="svelte-ws7k11"&&(Q.innerHTML=is),ve=r(e),M(z.$$.fragment,e),Ge=r(e),q=m(e,"P",{"data-svelte-h":!0}),U(q)!=="svelte-vsrbmk"&&(q.innerHTML=rs),_e=r(e),M(G.$$.fragment,e),We=r(e),_=m(e,"DIV",{class:!0,"data-svelte-h":!0}),U(_)!=="svelte-k6zsrl"&&(_.innerHTML=ps),Ve=r(e),W=m(e,"DIV",{class:!0,"data-svelte-h":!0}),U(W)!=="svelte-8ii4fv"&&(W.innerHTML=os),ke=r(e),D=m(e,"P",{"data-svelte-h":!0}),U(D)!=="svelte-95mtrv"&&(D.textContent=us),Se=r(e),M(Y.$$.fragment,e),xe=r(e),P=m(e,"P",{"data-svelte-h":!0}),U(P)!=="svelte-wsrnxa"&&(P.innerHTML=cs),Ce=r(e),M(V.$$.fragment,e),Ie=r(e),M(A.$$.fragment,e),Re=r(e),K=m(e,"P",{"data-svelte-h":!0}),U(K)!=="svelte-cu42wu"&&(K.textContent=ds),Be=r(e),M(O.$$.fragment,e),Xe=r(e),M(ee.$$.fragment,e),Ee=r(e),se=m(e,"P",{"data-svelte-h":!0}),U(se)!=="svelte-x7ex14"&&(se.innerHTML=fs),Fe=r(e),le=m(e,"P",{"data-svelte-h":!0}),U(le)!=="svelte-mll1yp"&&(le.innerHTML=ms),He=r(e),M(te.$$.fragment,e),Ne=r(e),ae=m(e,"P",{"data-svelte-h":!0}),U(ae)!=="svelte-1r8gmv0"&&(ae.innerHTML=hs),Le=r(e),M(ne.$$.fragment,e),Qe=r(e),ie=m(e,"P",{"data-svelte-h":!0}),U(ie)!=="svelte-vb7m03"&&(ie.innerHTML=Ms),ze=r(e),M(re.$$.fragment,e),qe=r(e),pe=m(e,"P",{"data-svelte-h":!0}),U(pe)!=="svelte-wirp0s"&&(pe.innerHTML=ys),De=r(e),M(oe.$$.fragment,e),Ye=r(e),ue=m(e,"P",{"data-svelte-h":!0}),U(ue)!=="svelte-dbifto"&&(ue.innerHTML=bs),Pe=r(e),M(ce.$$.fragment,e),Ae=r(e),fe=m(e,"P",{}),Us(fe).forEach(l),this.h()},h(){Oe(n,"name","hf:doc:metadata"),Oe(n,"content",Xs),Oe(_,"class","flex gap-4"),Oe(W,"class","flex gap-4")},m(e,s){vs(document.head,n),t(e,Z,s),t(e,u,s),t(e,o,s),y(c,e,s),t(e,a,s),y(d,e,s),t(e,v,s),t(e,p,s),t(e,k,s),t(e,j,s),t(e,S,s),y(x,e,s),t(e,Me,s),t(e,C,s),t(e,ye,s),y(I,e,s),t(e,be,s),y(R,e,s),t(e,we,s),t(e,B,s),t(e,ge,s),t(e,X,s),t(e,Ue,s),y(E,e,s),t(e,Ze,s),t(e,F,s),t(e,Je,s),y(H,e,s),t(e,Te,s),y(N,e,s),t(e,$e,s),t(e,L,s),t(e,je,s),t(e,Q,s),t(e,ve,s),y(z,e,s),t(e,Ge,s),t(e,q,s),t(e,_e,s),y(G,e,s),t(e,We,s),t(e,_,s),t(e,Ve,s),t(e,W,s),t(e,ke,s),t(e,D,s),t(e,Se,s),y(Y,e,s),t(e,xe,s),t(e,P,s),t(e,Ce,s),y(V,e,s),t(e,Ie,s),y(A,e,s),t(e,Re,s),t(e,K,s),t(e,Be,s),y(O,e,s),t(e,Xe,s),y(ee,e,s),t(e,Ee,s),t(e,se,s),t(e,Fe,s),t(e,le,s),t(e,He,s),y(te,e,s),t(e,Ne,s),t(e,ae,s),t(e,Le,s),y(ne,e,s),t(e,Qe,s),t(e,ie,s),t(e,ze,s),y(re,e,s),t(e,qe,s),t(e,pe,s),t(e,De,s),y(oe,e,s),t(e,Ye,s),t(e,ue,s),t(e,Pe,s),y(ce,e,s),t(e,Ae,s),t(e,fe,s),Ke=!0},p(e,[s]){const ws={};s&2&&(ws.$$scope={dirty:s,ctx:e}),G.$set(ws);const gs={};s&2&&(gs.$$scope={dirty:s,ctx:e}),V.$set(gs)},i(e){Ke||(b(c.$$.fragment,e),b(d.$$.fragment,e),b(x.$$.fragment,e),b(I.$$.fragment,e),b(R.$$.fragment,e),b(E.$$.fragment,e),b(H.$$.fragment,e),b(N.$$.fragment,e),b(z.$$.fragment,e),b(G.$$.fragment,e),b(Y.$$.fragment,e),b(V.$$.fragment,e),b(A.$$.fragment,e),b(O.$$.fragment,e),b(ee.$$.fragment,e),b(te.$$.fragment,e),b(ne.$$.fragment,e),b(re.$$.fragment,e),b(oe.$$.fragment,e),b(ce.$$.fragment,e),Ke=!0)},o(e){w(c.$$.fragment,e),w(d.$$.fragment,e),w(x.$$.fragment,e),w(I.$$.fragment,e),w(R.$$.fragment,e),w(E.$$.fragment,e),w(H.$$.fragment,e),w(N.$$.fragment,e),w(z.$$.fragment,e),w(G.$$.fragment,e),w(Y.$$.fragment,e),w(V.$$.fragment,e),w(A.$$.fragment,e),w(O.$$.fragment,e),w(ee.$$.fragment,e),w(te.$$.fragment,e),w(ne.$$.fragment,e),w(re.$$.fragment,e),w(oe.$$.fragment,e),w(ce.$$.fragment,e),Ke=!1},d(e){e&&(l(Z),l(u),l(o),l(a),l(v),l(p),l(k),l(j),l(S),l(Me),l(C),l(ye),l(be),l(we),l(B),l(ge),l(X),l(Ue),l(Ze),l(F),l(Je),l(Te),l($e),l(L),l(je),l(Q),l(ve),l(Ge),l(q),l(_e),l(We),l(_),l(Ve),l(W),l(ke),l(D),l(Se),l(xe),l(P),l(Ce),l(Ie),l(Re),l(K),l(Be),l(Xe),l(Ee),l(se),l(Fe),l(le),l(He),l(Ne),l(ae),l(Le),l(Qe),l(ie),l(ze),l(qe),l(pe),l(De),l(Ye),l(ue),l(Pe),l(Ae),l(fe)),l(n),g(c,e),g(d,e),g(x,e),g(I,e),g(R,e),g(E,e),g(H,e),g(N,e),g(z,e),g(G,e),g(Y,e),g(V,e),g(A,e),g(O,e),g(ee,e),g(te,e),g(ne,e),g(re,e),g(oe,e),g(ce,e)}}}const Xs='{"title":"Load schedulers and models","local":"load-schedulers-and-models","sections":[{"title":"Load a scheduler","local":"load-a-scheduler","sections":[],"depth":2},{"title":"Compare schedulers","local":"compare-schedulers","sections":[{"title":"Flax schedulers","local":"flax-schedulers","sections":[],"depth":3}],"depth":2},{"title":"Models","local":"models","sections":[],"depth":2}],"depth":1}';function Es($){return Js(()=>{new URLSearchParams(window.location.search).get("fw")}),[]}class Ds extends Ts{constructor(n){super(),$s(this,n,Es,Bs,Zs,{})}}export{Ds as component};