Buckets:

hf-doc-build
/

doc

Files
xet
hf-doc-build/doc / diffusers /main /fr /quantization /overview.html
mishig
1 day ago
download
raw
35.3 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Premiers pas&quot;,&quot;local&quot;:&quot;getting-started&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Quantification au niveau du pipeline&quot;,&quot;local&quot;:&quot;pipeline-level-quantization&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Quantification de base&quot;,&quot;local&quot;:&quot;basic-quantization&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Quantification avancée&quot;,&quot;local&quot;:&quot;advanced-quantization&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Chargement des points de contrôle pré-quantifiés&quot;,&quot;local&quot;:&quot;loading-prequantized-checkpoints&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Ressources&quot;,&quot;local&quot;:&quot;resources&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2}],&quot;depth&quot;:1}"/>
		<link href="/docs/diffusers/main/en/_app/immutable/entry/start.D3LnKZ4R.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/BF9UNEM9.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DPnyVmXs.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/entry/app.C0rBoMxR.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/DSGUT_OW.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/C045MWWx.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.DTZsP_3B.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/IaDvMxLQ.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/nodes/313.C7ff02l6.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/bKdwVYpU.js" rel="modulepreload">
		<!--cubdu--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Getting started&quot;,&quot;local&quot;:&quot;getting-started&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Pipeline-level quantization&quot;,&quot;local&quot;:&quot;pipeline-level-quantization&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Basic quantization&quot;,&quot;local&quot;:&quot;basic-quantization&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3},{&quot;title&quot;:&quot;Advanced quantization&quot;,&quot;local&quot;:&quot;advanced-quantization&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Loading prequantized checkpoints&quot;,&quot;local&quot;:&quot;loading-prequantized-checkpoints&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Resources&quot;,&quot;local&quot;:&quot;resources&quot;,&quot;sections&quot;:[],&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="getting-started" 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="#getting-started"><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><!---->Premiers pas<!----></span></h1><!--]--><!----> <p>La quantisation consiste à représenter les données avec moins de bits tout en cherchant à préserver la précision des données d’origine. Cela implique souvent de convertir un type de données pour représenter les mêmes informations avec moins de bits. Par exemple, si les poids de votre modèle sont stockés en nombres flottants sur 32 bits et qu’ils sont quantifiés en nombres flottants sur 16 bits, la taille du modèle est divisée par deux, ce qui facilite son stockage et réduit l’utilisation de la mémoire. Une précision plus faible peut également accélérer l’inférence, car les calculs avec moins de bits prennent moins de temps.</p> <p>Diffusers prend en charge plusieurs backends de quantification pour rendre les grands modèles de diffusion comme <a href="../api/pipelines/flux">Flux</a> plus accessibles. Ce guide montre comment utiliser la classe <a href="/docs/diffusers/main/fr/api/quantization#diffusers.PipelineQuantizationConfig">PipelineQuantizationConfig</a> pour quantifier un pipeline lors de son initialisation à partir d'un checkpoint pré-entraîné ou non quantifié.</p> <!--[1--><h2 class="relative group"><a id="pipeline-level-quantization" 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="#pipeline-level-quantization"><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><!---->Quantification au niveau du pipeline<!----></span></h2><!--]--><!----> <p>Il existe deux façons d’utiliser <a href="/docs/diffusers/main/fr/api/quantization#diffusers.PipelineQuantizationConfig">PipelineQuantizationConfig</a> en fonction du niveau de personnalisation que vous souhaitez appliquer à la configuration de quantification.</p> <ul><li>pour les cas d'utilisation de base, définissez les arguments <code>quant_backend</code>, <code>quant_kwargs</code> et <code>components_to_quantize</code></li> <li>pour un contrôle granulaire de la quantification, définissez un <code>quant_mapping</code> qui fournit la configuration de quantification pour les composants individuels du modèle</li></ul> <!--[2--><h3 class="relative group"><a id="basic-quantization" 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="#basic-quantization"><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><!---->Quantification de base<!----></span></h3><!--]--><!----> <p>Initialisez <a href="/docs/diffusers/main/fr/api/quantization#diffusers.PipelineQuantizationConfig">PipelineQuantizationConfig</a> avec les paramètres suivants.</p> <ul><li><code>quant_backend</code> spécifie le backend de quantification à utiliser. Les backends actuellement pris en charge incluent : <code>bitsandbytes_4bit</code>, <code>bitsandbytes_8bit</code>, <code>gguf</code>, <code>quanto</code> et <code>torchao</code>.</li> <li><code>quant_kwargs</code> spécifie les arguments de quantification à utiliser.</li></ul> <blockquote class="tip"><p>Ces arguments <code>quant_kwargs</code> diffèrent selon chaque backend. Consultez la documentation de l'<a href="../api/quantization">API de quantification</a> pour afficher les arguments de chaque backend.</p></blockquote> <ul><li><p><code>components_to_quantize</code> spécifie le ou les composants du pipeline à quantiser. En général, vous devez quantiser les composants les plus gourmands en calcul, comme le transformer. L’encodeur de texte est un autre composant à prendre en compte pour la quantisation si le pipeline en comporte plus d’un, comme <a href="/docs/diffusers/main/fr/api/pipelines/flux#diffusers.FluxPipeline">FluxPipeline</a>. L’exemple ci-dessous quantifie l’encodeur de texte T5 dans <a href="/docs/diffusers/main/fr/api/pipelines/flux#diffusers.FluxPipeline">FluxPipeline</a> tout en conservant le modèle CLIP intact.</p> <p><code>components_to_quantize</code> accepte soit une liste pour plusieurs modèles, soit une chaîne de caractères pour un modèle unique.</p></li></ul> <p>L'exemple ci-dessous charge le backend bitsandbytes avec les arguments suivants issus de <a href="/docs/diffusers/main/fr/api/quantization#diffusers.BitsAndBytesConfig">BitsAndBytesConfig</a>, <code>load_in_4bit</code>, <code>bnb_4bit_quant_type</code> et <code>bnb_4bit_compute_dtype</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> Copié</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
<span class="hljs-keyword">from</span> diffusers.quantizers <span class="hljs-keyword">import</span> PipelineQuantizationConfig

pipeline_quant_config = PipelineQuantizationConfig(
    quant_backend=<span class="hljs-string">&quot;bitsandbytes_4bit&quot;</span>,
    quant_kwargs={<span class="hljs-string">&quot;load_in_4bit&quot;</span>: <span class="hljs-literal">True</span>, <span class="hljs-string">&quot;bnb_4bit_quant_type&quot;</span>: <span class="hljs-string">&quot;nf4&quot;</span>, <span class="hljs-string">&quot;bnb_4bit_compute_dtype&quot;</span>: torch.bfloat16},
    components_to_quantize=[<span class="hljs-string">&quot;transformer&quot;</span>, <span class="hljs-string">&quot;text_encoder_2&quot;</span>],
)<!----></pre></div><!----> <p>Passez <code>pipeline_quant_config</code> à <a href="/docs/diffusers/main/fr/api/pipelines/overview#diffusers.DiffusionPipeline.from_pretrained">from_pretrained()</a> pour quantifier le 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> Copié</div><!----></button><!----></div> <pre class="language-py "><!---->pipe = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;black-forest-labs/FLUX.1-dev&quot;</span>,
    quantization_config=pipeline_quant_config,
    dtype=torch.bfloat16,
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

image = pipe(<span class="hljs-string">&quot;photo of a cute dog&quot;</span>).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <!--[2--><h3 class="relative group"><a id="advanced-quantization" 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="#advanced-quantization"><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><!---->Quantification avancée<!----></span></h3><!--]--><!----> <p>L'argument <code>quant_mapping</code> offre davantage d'options pour quantiser chaque composant individuel d'un pipeline, comme la combinaison de différents backends de quantification.</p> <p>Initialisez <a href="/docs/diffusers/main/fr/api/quantization#diffusers.PipelineQuantizationConfig">PipelineQuantizationConfig</a> et passez-lui un <code>quant_mapping</code>. Le <code>quant_mapping</code> vous permet de spécifier les options de quantisation pour chaque composant du pipeline, tel que le transformer et l'encodeur de texte.</p> <p>L'exemple ci-dessous utilise deux backends de quantification, <a href="/docs/diffusers/main/fr/api/quantization#diffusers.QuantoConfig">QuantoConfig</a> et <a href="https://huggingface.co/docs/transformers/main/en/main_classes/quantization#transformers.BitsAndBytesConfig" rel="nofollow">transformers.BitsAndBytesConfig</a>, pour le transformer et l'encodeur de texte.</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> Copié</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
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> BitsAndBytesConfig <span class="hljs-keyword">as</span> DiffusersBitsAndBytesConfig
<span class="hljs-keyword">from</span> diffusers.quantizers.quantization_config <span class="hljs-keyword">import</span> QuantoConfig
<span class="hljs-keyword">from</span> diffusers.quantizers <span class="hljs-keyword">import</span> PipelineQuantizationConfig
<span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> BitsAndBytesConfig <span class="hljs-keyword">as</span> TransformersBitsAndBytesConfig

pipeline_quant_config = PipelineQuantizationConfig(
    quant_mapping={
        <span class="hljs-string">&quot;transformer&quot;</span>: QuantoConfig(weights_dtype=<span class="hljs-string">&quot;int8&quot;</span>),
        <span class="hljs-string">&quot;text_encoder_2&quot;</span>: TransformersBitsAndBytesConfig(
            load_in_4bit=<span class="hljs-literal">True</span>, compute_dtype=torch.bfloat16
        ),
    }
)<!----></pre></div><!----> <p>Il existe un backend bitsandbytes distinct dans <a href="https://huggingface.co/docs/transformers/main_classes/quantization#transformers.BitsAndBytesConfig" rel="nofollow">Transformers</a>. Vous devez importer et utiliser <a href="https://huggingface.co/docs/transformers/main/en/main_classes/quantization#transformers.BitsAndBytesConfig" rel="nofollow">transformers.BitsAndBytesConfig</a> pour les composants provenant de Transformers. Par exemple, <code>text_encoder_2</code> dans <a href="/docs/diffusers/main/fr/api/pipelines/flux#diffusers.FluxPipeline">FluxPipeline</a> est un <a href="https://huggingface.co/docs/transformers/main/en/model_doc/t5#transformers.T5EncoderModel" rel="nofollow">T5EncoderModel</a> de Transformers, il faut donc utiliser <a href="https://huggingface.co/docs/transformers/main/en/main_classes/quantization#transformers.BitsAndBytesConfig" rel="nofollow">transformers.BitsAndBytesConfig</a> au lieu de <a href="/docs/diffusers/main/fr/api/quantization#diffusers.BitsAndBytesConfig">diffusers.BitsAndBytesConfig</a>.</p> <blockquote class="tip"><p>Utilisez la méthode de <a href="#basic-quantization">quantification de base</a> ci-dessus si vous ne souhaitez pas gérer ces imports distincts ou si vous n’êtes pas sûr de l’origine de chaque composant du pipeline.</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> Copié</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
<span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> BitsAndBytesConfig <span class="hljs-keyword">as</span> DiffusersBitsAndBytesConfig
<span class="hljs-keyword">from</span> diffusers.quantizers <span class="hljs-keyword">import</span> PipelineQuantizationConfig
<span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> BitsAndBytesConfig <span class="hljs-keyword">as</span> TransformersBitsAndBytesConfig

pipeline_quant_config = PipelineQuantizationConfig(
    quant_mapping={
        <span class="hljs-string">&quot;transformer&quot;</span>: DiffusersBitsAndBytesConfig(load_in_4bit=<span class="hljs-literal">True</span>, bnb_4bit_compute_dtype=torch.bfloat16),
        <span class="hljs-string">&quot;text_encoder_2&quot;</span>: TransformersBitsAndBytesConfig(
            load_in_4bit=<span class="hljs-literal">True</span>, compute_dtype=torch.bfloat16
        ),
    }
)<!----></pre></div><!----> <p>Passez <code>pipeline_quant_config</code> à <a href="/docs/diffusers/main/fr/api/pipelines/overview#diffusers.DiffusionPipeline.from_pretrained">from_pretrained()</a> pour quantifier le 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> Copié</div><!----></button><!----></div> <pre class="language-py "><!---->pipe = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;black-forest-labs/FLUX.1-dev&quot;</span>,
    quantization_config=pipeline_quant_config,
    dtype=torch.bfloat16,
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

image = pipe(<span class="hljs-string">&quot;photo of a cute dog&quot;</span>).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="loading-prequantized-checkpoints" 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="#loading-prequantized-checkpoints"><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><!---->Chargement des points de contrôle pré-quantifiés<!----></span></h2><!--]--><!----> <p>Certains backends de quantisation prennent en charge le chargement de points de contrôle déjà quantisés et téléversés sur le Hub. Dans ce cas, la configuration de quantisation est stockée dans le <code>config.json</code> du modèle et est lue automatiquement — vous n’avez pas besoin de créer un <a href="/docs/diffusers/main/fr/api/quantization#diffusers.PipelineQuantizationConfig">PipelineQuantizationConfig</a> ni de passer un argument <code>quantization_config</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> Copié</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

pipe = DiffusionPipeline.from_pretrained(
    <span class="hljs-string">&quot;rootonchair/ERNIE-Image-Turbo-nunchaku-lite-nvfp4&quot;</span>,
    dtype=torch.bfloat16,
).to(<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span><!----></pre></div><!----> <p>Les backends suivants prennent en charge le chargement de points de contrôle préquantifiés sans configuration supplémentaire.</p> <table><thead><tr><th>Backend</th><th>Notes</th></tr></thead><tbody><tr><td><a href="./bitsandbytes">bitsandbytes</a></td><td>La configuration est enregistrée dans <code>config.json</code> ; aucun argument supplémentaire n'est nécessaire. Prend en charge la quantification à la volée ainsi que le chargement de points de contrôle pré-quantifiés.</td></tr><tr><td><a href="./torchao">torchao</a></td><td>Identique à ce qui précède.</td></tr><tr><td><a href="./gguf">GGUF</a></td><td>Utilise <code>from_single_file</code> avec les classes de modèle ; le chargement au niveau du pipeline n'est pas pris en charge.</td></tr><tr><td><a href="./bfl">BFL</a></td><td>Charge les points de contrôle monofichier <code>fp8r</code> et <code>nvfp4</code> de Black Forest Labs avec <code>from_single_file</code> ; le chargement au niveau du pipeline n'est pas pris en charge.</td></tr><tr><td><a href="./autoround">AutoRound</a></td><td>Seul le chargement est pris en charge ; quantifiez d'abord avec l'interface en ligne de commande AutoRound ou l'API Python.</td></tr><tr><td><a href="./nunchaku">Nunchaku Lite</a></td><td>La configuration est enregistrée dans <code>config.json</code> ; le package <code>kernels</code> est requis. Seul le chargement est pris en charge.</td></tr><tr><td><a href="./modelopt">ModelOpt</a></td><td>Seul le chargement de modèles pré-quantifiés est pris en charge.</td></tr><tr><td><a href="./sdnq">SDNQ</a></td><td>Prend en charge la quantification à la volée et le chargement de modèles pré-quantifiés ; nécessite le package <code>sdnq</code>.</td></tr></tbody></table> <!--[1--><h2 class="relative group"><a id="resources" 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="#resources"><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><!---->Ressources<!----></span></h2><!--]--><!----> <p>Consultez les ressources ci-dessous pour en savoir plus sur la quantification.</p> <ul><li><p>Si vous débutez en quantification, nous vous recommandons de consulter les cours suivants, conçus pour les débutants et réalisés en collaboration avec DeepLearning.AI.</p> <ul><li><a href="https://www.deeplearning.ai/short-courses/quantization-fundamentals-with-hugging-face/" rel="nofollow">Fondamentaux de la quantification avec Hugging Face</a></li> <li><a href="https://www.deeplearning.ai/short-courses/quantization-in-depth/" rel="nofollow">La quantification en profondeur</a></li></ul></li> <li><p>Consultez le <a href="https://huggingface.co/docs/transformers/main/en/quantization/contribute" rel="nofollow">guide pour contribuer avec une nouvelle méthode de quantification</a> si vous souhaitez ajouter une nouvelle méthode de quantification.</p></li> <li><p>L'<a href="https://huggingface.co/docs/transformers/quantization/overview#when-to-use-what" rel="nofollow">aperçu</a> de la quantification dans Transformers présente les avantages et les inconvénients des différents backends de quantification.</p></li> <li><p>Lisez le billet de blog <a href="https://huggingface.co/blog/diffusers-quantization" rel="nofollow">Exploring Quantization Backends in Diffusers</a> pour une brève introduction à chaque backend de quantification, à la manière de choisir un backend et à la combinaison de la quantification avec d’autres optimisations de mémoire.</p></li></ul> <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/quantization/overview.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_g37pnb = {
						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.D3LnKZ4R.js"),
						import("/docs/diffusers/main/en/_app/immutable/entry/app.C0rBoMxR.js")
					]).then(([kit, app]) => {
						kit.start(app, element, {
							node_ids: [0, 313],
							data: [null,null],
							form: null,
							error: null
						});
					});
				}
			</script>
		
Xet Storage Details

Size:: 35.3 kB
Xet hash:: 754dcf8b1a0967cac31e84bace1406dcdd901996bda7b33a2157e0501e7d8908
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.