diffusers/pr_13832/en/_app/immutable/nodes/332.423a1145.js

import{s as Wl,n as Cl,o as Xl}from"../chunks/scheduler.53228c21.js";import{S as xl,i as vl,e as i,s as a,c as y,h as kl,a as M,d as s,b as n,f as Il,g as c,j as p,k as H,l as Yl,m as t,n as m,t as J,o as r,p as h}from"../chunks/index.cac5d66a.js";import{C as Ql}from"../chunks/CopyLLMTxtMenu.efae84b2.js";import{C as E}from"../chunks/CodeBlock.606cbaf4.js";import{H as gl,E as Sl}from"../chunks/MermaidChart.svelte_svelte_type_style_lang.de76e98b.js";function Hl(Jl){let o,R,N,z,b,F,j,$,u,rl="Batch inference processes multiple prompts at a time to increase throughput. It is more efficient because processing multiple prompts at once maximizes GPU usage versus processing a single prompt and underutilizing the GPU.",_,G,hl="The downside is increased latency because you must wait for the entire batch to complete, and more GPU memory is required for large batches.",A,B,wl="For text-to-image, pass a list of prompts to the pipeline and for image-to-image, pass a list of images and prompts to the pipeline. The example below demonstrates batched text-to-image inference.",L,Z,q,w,dl='<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/batch-inference.png"/>',D,f,Tl="To generate multiple variations of one prompt, use the <code>num_images_per_prompt</code> argument.",P,I,K,d,Ul='<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/batch-inference-2.png"/>',O,g,bl="Combine both approaches to generate different variations of different prompts.",ll,W,el,T,jl='<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/batch-inference-3.png"/>',sl,C,tl,X,ul='Enable reproducible batch generation by passing a list of <a href="https://pytorch.org/docs/stable/generated/torch.Generator.html" rel="nofollow">Generator’s</a> to the pipeline and tie each <code>Generator</code> to a seed to reuse it.',al,U,Gl='<p>Refer to the <a href="./reusing_seeds">Reproducibility</a> docs to learn more about deterministic algorithms and the <code>Generator</code> object.</p>',nl,x,Bl="Use a list comprehension to iterate over the batch size specified in <code>range()</code> to create a unique <code>Generator</code> object for each image in the batch. Don’t multiply the <code>Generator</code> by the batch size because that only creates one <code>Generator</code> object that is used sequentially for each image in the batch.",il,v,Ml,k,Zl="Pass the <code>generator</code> to the pipeline.",pl,Y,ol,Q,fl="You can use this to select an image associated with a seed and iteratively improve on it by crafting a more detailed prompt.",yl,S,cl,V,ml;return b=new Ql({props:{containerStyle:"float: right; margin-left: 10px; display: inline-flex; position: relative; z-index: 10;"}}),j=new gl({props:{title:"Batch inference",local:"batch-inference",headingTag:"h1"}}),Z=new E({props:{code:"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",highlighted:`<span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">import</span> matplotlib.pyplot <span class="hljs-keyword">as</span> plt
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

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

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

images = pipeline(
    prompt=prompts,
).images

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

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

plt.tight_layout()
plt.show()`,lang:"py",wrap:!1}}),I=new E({props:{code:"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",highlighted:`<span class="hljs-keyword">import</span> torch
<span class="hljs-keyword">import</span> matplotlib.pyplot <span class="hljs-keyword">as</span> plt
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline

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

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

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

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

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

plt.tight_layout()
plt.show()`,lang:"py",wrap:!1}}),W=new E({props:{code:"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",highlighted:`images = pipeline(
    prompt=prompts,
    num_images_per_prompt=<span class="hljs-number">2</span>,
).images

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

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

plt.tight_layout()
plt.show()`,lang:"py",wrap:!1}}),C=new gl({props:{title:"Deterministic generation",local:"deterministic-generation",headingTag:"h2"}}),v=new E({props:{code:"Z2VuZXJhdG9yJTIwJTNEJTIwJTVCdG9yY2guR2VuZXJhdG9yKGRldmljZSUzRCUyMmN1ZGElMjIpLm1hbnVhbF9zZWVkKDApJTVEJTIwKiUyMDM=",highlighted:'generator = [torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">0</span>)] * <span class="hljs-number">3</span>',lang:"py",wrap:!1}}),Y=new E({props:{code:"aW1wb3J0JTIwdG9yY2glMEFmcm9tJTIwZGlmZnVzZXJzJTIwaW1wb3J0JTIwRGlmZnVzaW9uUGlwZWxpbmUlMEElMEFwaXBlbGluZSUyMCUzRCUyMERpZmZ1c2lvblBpcGVsaW5lLmZyb21fcHJldHJhaW5lZCglMEElMjAlMjAlMjAlMjAlMjJzdGFiaWxpdHlhaSUyRnN0YWJsZS1kaWZmdXNpb24teGwtYmFzZS0xLjAlMjIlMkMlMEElMjAlMjAlMjAlMjB0b3JjaF9kdHlwZSUzRHRvcmNoLmZsb2F0MTYlMkMlMEElMjAlMjAlMjAlMjBkZXZpY2VfbWFwJTNEJTIyY3VkYSUyMiUwQSklMEElMEFnZW5lcmF0b3IlMjAlM0QlMjAlNUJ0b3JjaC5HZW5lcmF0b3IoZGV2aWNlJTNEJTIyY3VkYSUyMikubWFudWFsX3NlZWQoaSklMjBmb3IlMjBpJTIwaW4lMjByYW5nZSgzKSU1RCUwQXByb21wdHMlMjAlM0QlMjAlNUIlMEElMjAlMjAlMjAlMjAlMjJDaW5lbWF0aWMlMjBzaG90JTIwb2YlMjBhJTIwY296eSUyMGNvZmZlZSUyMHNob3AlMjBpbnRlcmlvciUyQyUyMHdhcm0lMjBwYXN0ZWwlMjBsaWdodCUyMHN0cmVhbWluZyUyMHRocm91Z2glMjBhJTIwd2luZG93JTIwd2hlcmUlMjBhJTIwY2F0JTIwcmVzdHMuJTIwU2hhbGxvdyUyMGRlcHRoJTIwb2YlMjBmaWVsZCUyQyUyMGdsb3dpbmclMjBjdXBzJTIwaW4lMjBzb2Z0JTIwZm9jdXMlMkMlMjBkcmVhbXklMjBsb2ZpLWluc3BpcmVkJTIwbW9vZCUyQyUyMG5vc3RhbGdpYyUyMHRvbmVzJTJDJTIwZnJhbWVkJTIwbGlrZSUyMGElMjBxdWlldCUyMGZpbG0lMjBzY2VuZS4lMjIlMkMlMEElMjAlMjAlMjAlMjAlMjJQb2xhcm9pZC1zdHlsZSUyMHBob3RvZ3JhcGglMjBvZiUyMGElMjBjb3p5JTIwY29mZmVlJTIwc2hvcCUyMGludGVyaW9yJTJDJTIwYmF0aGVkJTIwaW4lMjB3YXJtJTIwcGFzdGVsJTIwbGlnaHQuJTIwQSUyMGNhdCUyMHNpdHMlMjBvbiUyMHRoZSUyMHdpbmRvd3NpbGwlMjBuZWFyJTIwc3RlYW1pbmclMjBtdWdzLiUyMFNvZnQlMkMlMjBzbGlnaHRseSUyMGZhZGVkJTIwdG9uZXMlMjBhbmQlMjBkcmVhbXklMjBibHVyJTIwZXZva2UlMjBub3N0YWxnaWElMkMlMjBhJTIwbG9maSUyMG1vb2QlMkMlMjBhbmQlMjB0aGUlMjBpbnRpbWF0ZSUyQyUyMGltcGVyZmVjdCUyMGNoYXJtJTIwb2YlMjBpbnN0YW50JTIwZmlsbS4lMjIlMkMlMEElMjAlMjAlMjAlMjAlMjJTb2Z0JTIwd2F0ZXJjb2xvciUyMGlsbHVzdHJhdGlvbiUyMG9mJTIwYSUyMGNvenklMjBjb2ZmZWUlMjBzaG9wJTIwaW50ZXJpb3IlMkMlMjBwYXN0ZWwlMjB3YXNoZXMlMjBvZiUyMGNvbG9yJTIwZmlsbGluZyUyMHRoZSUyMHNwYWNlLiUyMEElMjBjYXQlMjByZXN0cyUyMHBlYWNlZnVsbHklMjBvbiUyMHRoZSUyMHdpbmRvd3NpbGwlMjBhcyUyMHdhcm0lMjBsaWdodCUyMGdsb3dzJTIwdGhyb3VnaC4lMjBHZW50bGUlMjBicnVzaHN0cm9rZXMlMjBjcmVhdGUlMjBhJTIwZHJlYW15JTJDJTIwbG9maS1pbnNwaXJlZCUyMGF0bW9zcGhlcmUlMjB3aXRoJTIwd2hpbXNpY2FsJTIwdGV4dHVyZXMlMjBhbmQlMjBub3N0YWxnaWMlMjBjYWxtLiUyMiUyQyUwQSUyMCUyMCUyMCUyMCUyMklzb21ldHJpYyUyMHBpeGVsLWFydCUyMGlsbHVzdHJhdGlvbiUyMG9mJTIwYSUyMGNvenklMjBjb2ZmZWUlMjBzaG9wJTIwaW50ZXJpb3IlMjBpbiUyMGRldGFpbGVkJTIwOC1iaXQlMjBzdHlsZS4lMjBXYXJtJTIwcGFzdGVsJTIwbGlnaHQlMjBmaWxscyUyMHRoZSUyMHNwYWNlJTIwYXMlMjBhJTIwY2F0JTIwcmVzdHMlMjBvbiUyMHRoZSUyMHdpbmRvd3NpbGwuJTIwQmxvY2t5JTIwZnVybml0dXJlJTIwYW5kJTIwdGlueSUyMG11Z3MlMjBhZGQlMjBjaGFybSUyQyUyMGxvdy1yZXMlMjByZXRybyUyMGdyYXBoaWNzJTIwZW5oYW5jZSUyMHRoZSUyMG5vc3RhbGdpYyUyQyUyMGxvZmktaW5zcGlyZWQlMjBnYW1lJTIwYWVzdGhldGljLiUyMiUwQSU1RCUwQSUwQWltYWdlcyUyMCUzRCUyMHBpcGVsaW5lKCUwQSUyMCUyMCUyMCUyMHByb21wdCUzRHByb21wdHMlMkMlMEElMjAlMjAlMjAlMjBnZW5lcmF0b3IlM0RnZW5lcmF0b3IlMEEpLmltYWdlcyUwQSUwQWZpZyUyQyUyMGF4ZXMlMjAlM0QlMjBwbHQuc3VicGxvdHMoMiUyQyUyMDIlMkMlMjBmaWdzaXplJTNEKDEyJTJDJTIwMTIpKSUwQWF4ZXMlMjAlM0QlMjBheGVzLmZsYXR0ZW4oKSUwQSUwQWZvciUyMGklMkMlMjBpbWFnZSUyMGluJTIwZW51bWVyYXRlKGltYWdlcyklM0ElMEElMjAlMjAlMjAlMjBheGVzJTVCaSU1RC5pbXNob3coaW1hZ2UpJTBBJTIwJTIwJTIwJTIwYXhlcyU1QmklNUQuc2V0X3RpdGxlKGYlMjJJbWFnZSUyMCU3QmklMkIxJTdEJTIyKSUwQSUyMCUyMCUyMCUyMGF4ZXMlNUJpJTVELmF4aXMoJ29mZicpJTBBJTBBcGx0LnRpZ2h0X2xheW91dCgpJTBBcGx0LnNob3coKQ==",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;stabilityai/stable-diffusion-xl-base-1.0&quot;</span>,
    torch_dtype=torch.float16,
    device_map=<span class="hljs-string">&quot;cuda&quot;</span>
)

generator = [torch.Generator(device=<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(i) <span class="hljs-keyword">for</span> i <span class="hljs-keyword">in</span> <span class="hljs-built_in">range</span>(<span class="hljs-number">3</span>)]
prompts = [
    <span class="hljs-string">&quot;Cinematic shot of a cozy coffee shop interior, warm pastel light streaming through a window where a cat rests. Shallow depth of field, glowing cups in soft focus, dreamy lofi-inspired mood, nostalgic tones, framed like a quiet film scene.&quot;</span>,
    <span class="hljs-string">&quot;Polaroid-style photograph of a cozy coffee shop interior, bathed in warm pastel light. A cat sits on the windowsill near steaming mugs. Soft, slightly faded tones and dreamy blur evoke nostalgia, a lofi mood, and the intimate, imperfect charm of instant film.&quot;</span>,
    <span class="hljs-string">&quot;Soft watercolor illustration of a cozy coffee shop interior, pastel washes of color filling the space. A cat rests peacefully on the windowsill as warm light glows through. Gentle brushstrokes create a dreamy, lofi-inspired atmosphere with whimsical textures and nostalgic calm.&quot;</span>,
    <span class="hljs-string">&quot;Isometric pixel-art illustration of a cozy coffee shop interior in detailed 8-bit style. Warm pastel light fills the space as a cat rests on the windowsill. Blocky furniture and tiny mugs add charm, low-res retro graphics enhance the nostalgic, lofi-inspired game aesthetic.&quot;</span>
]

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

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

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

plt.tight_layout()
plt.show()`,lang:"py",wrap:!1}}),S=new Sl({props:{source:"https://github.com/huggingface/diffusers/blob/main/docs/source/en/using-diffusers/batched_inference.md"}}),{c(){o=i("meta"),R=a(),N=i("p"),z=a(),y(b.$$.fragment),F=a(),y(j.$$.fragment),$=a(),u=i("p"),u.textContent=rl,_=a(),G=i("p"),G.textContent=hl,A=a(),B=i("p"),B.textContent=wl,L=a(),y(Z.$$.fragment),q=a(),w=i("div"),w.innerHTML=dl,D=a(),f=i("p"),f.innerHTML=Tl,P=a(),y(I.$$.fragment),K=a(),d=i("div"),d.innerHTML=Ul,O=a(),g=i("p"),g.textContent=bl,ll=a(),y(W.$$.fragment),el=a(),T=i("div"),T.innerHTML=jl,sl=a(),y(C.$$.fragment),tl=a(),X=i("p"),X.innerHTML=ul,al=a(),U=i("blockquote"),U.innerHTML=Gl,nl=a(),x=i("p"),x.innerHTML=Bl,il=a(),y(v.$$.fragment),Ml=a(),k=i("p"),k.innerHTML=Zl,pl=a(),y(Y.$$.fragment),ol=a(),Q=i("p"),Q.textContent=fl,yl=a(),y(S.$$.fragment),cl=a(),V=i("p"),this.h()},l(l){const e=kl("svelte-u9bgzb",document.head);o=M(e,"META",{name:!0,content:!0}),e.forEach(s),R=n(l),N=M(l,"P",{}),Il(N).forEach(s),z=n(l),c(b.$$.fragment,l),F=n(l),c(j.$$.fragment,l),$=n(l),u=M(l,"P",{"data-svelte-h":!0}),p(u)!=="svelte-1gcbm5z"&&(u.textContent=rl),_=n(l),G=M(l,"P",{"data-svelte-h":!0}),p(G)!=="svelte-19m5zxe"&&(G.textContent=hl),A=n(l),B=M(l,"P",{"data-svelte-h":!0}),p(B)!=="svelte-xbr0fq"&&(B.textContent=wl),L=n(l),c(Z.$$.fragment,l),q=n(l),w=M(l,"DIV",{class:!0,"data-svelte-h":!0}),p(w)!=="svelte-1nijj5t"&&(w.innerHTML=dl),D=n(l),f=M(l,"P",{"data-svelte-h":!0}),p(f)!=="svelte-1wlqa91"&&(f.innerHTML=Tl),P=n(l),c(I.$$.fragment,l),K=n(l),d=M(l,"DIV",{class:!0,"data-svelte-h":!0}),p(d)!=="svelte-7ab4sg"&&(d.innerHTML=Ul),O=n(l),g=M(l,"P",{"data-svelte-h":!0}),p(g)!=="svelte-1tfp3na"&&(g.textContent=bl),ll=n(l),c(W.$$.fragment,l),el=n(l),T=M(l,"DIV",{class:!0,"data-svelte-h":!0}),p(T)!=="svelte-w8ud0d"&&(T.innerHTML=jl),sl=n(l),c(C.$$.fragment,l),tl=n(l),X=M(l,"P",{"data-svelte-h":!0}),p(X)!=="svelte-agd9mv"&&(X.innerHTML=ul),al=n(l),U=M(l,"BLOCKQUOTE",{class:!0,"data-svelte-h":!0}),p(U)!=="svelte-14xuqs2"&&(U.innerHTML=Gl),nl=n(l),x=M(l,"P",{"data-svelte-h":!0}),p(x)!=="svelte-1qnmdlw"&&(x.innerHTML=Bl),il=n(l),c(v.$$.fragment,l),Ml=n(l),k=M(l,"P",{"data-svelte-h":!0}),p(k)!=="svelte-c466x0"&&(k.innerHTML=Zl),pl=n(l),c(Y.$$.fragment,l),ol=n(l),Q=M(l,"P",{"data-svelte-h":!0}),p(Q)!=="svelte-4k5zjx"&&(Q.textContent=fl),yl=n(l),c(S.$$.fragment,l),cl=n(l),V=M(l,"P",{}),Il(V).forEach(s),this.h()},h(){H(o,"name","hf:doc:metadata"),H(o,"content",Nl),H(w,"class","flex justify-center"),H(d,"class","flex justify-center"),H(T,"class","flex justify-center"),H(U,"class","tip")},m(l,e){Yl(document.head,o),t(l,R,e),t(l,N,e),t(l,z,e),m(b,l,e),t(l,F,e),m(j,l,e),t(l,$,e),t(l,u,e),t(l,_,e),t(l,G,e),t(l,A,e),t(l,B,e),t(l,L,e),m(Z,l,e),t(l,q,e),t(l,w,e),t(l,D,e),t(l,f,e),t(l,P,e),m(I,l,e),t(l,K,e),t(l,d,e),t(l,O,e),t(l,g,e),t(l,ll,e),m(W,l,e),t(l,el,e),t(l,T,e),t(l,sl,e),m(C,l,e),t(l,tl,e),t(l,X,e),t(l,al,e),t(l,U,e),t(l,nl,e),t(l,x,e),t(l,il,e),m(v,l,e),t(l,Ml,e),t(l,k,e),t(l,pl,e),m(Y,l,e),t(l,ol,e),t(l,Q,e),t(l,yl,e),m(S,l,e),t(l,cl,e),t(l,V,e),ml=!0},p:Cl,i(l){ml||(J(b.$$.fragment,l),J(j.$$.fragment,l),J(Z.$$.fragment,l),J(I.$$.fragment,l),J(W.$$.fragment,l),J(C.$$.fragment,l),J(v.$$.fragment,l),J(Y.$$.fragment,l),J(S.$$.fragment,l),ml=!0)},o(l){r(b.$$.fragment,l),r(j.$$.fragment,l),r(Z.$$.fragment,l),r(I.$$.fragment,l),r(W.$$.fragment,l),r(C.$$.fragment,l),r(v.$$.fragment,l),r(Y.$$.fragment,l),r(S.$$.fragment,l),ml=!1},d(l){l&&(s(R),s(N),s(z),s(F),s($),s(u),s(_),s(G),s(A),s(B),s(L),s(q),s(w),s(D),s(f),s(P),s(K),s(d),s(O),s(g),s(ll),s(el),s(T),s(sl),s(tl),s(X),s(al),s(U),s(nl),s(x),s(il),s(Ml),s(k),s(pl),s(ol),s(Q),s(yl),s(cl),s(V)),s(o),h(b,l),h(j,l),h(Z,l),h(I,l),h(W,l),h(C,l),h(v,l),h(Y,l),h(S,l)}}}const Nl='{"title":"Batch inference","local":"batch-inference","sections":[{"title":"Deterministic generation","local":"deterministic-generation","sections":[],"depth":2}],"depth":1}';function Vl(Jl){return Xl(()=>{new URLSearchParams(window.location.search).get("fw")}),[]}class _l extends xl{constructor(o){super(),vl(this,o,Vl,Hl,Wl,{})}}export{_l as component};