Buckets:

hf-doc-build
/

doc

Files
xet
hf-doc-build/doc / diffusers /main /ja /api /pipelines /pika_h3_flex.html
mishig
1 day ago
download
raw
114 kB
<meta charset="utf-8" /><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Pika H3 Flex&quot;,&quot;local&quot;:&quot;pika-h3-flex&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;読み込み中&quot;,&quot;local&quot;:&quot;loading&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;読み込んだMiniMax-H3の再利用&quot;,&quot;local&quot;:&quot;reusing-a-loaded-minimax-h3&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;MiniMax-H3との違い&quot;,&quot;local&quot;:&quot;what-differs-from-minimax-h3&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;生成&quot;,&quot;local&quot;:&quot;generation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexModularPipeline&quot;,&quot;local&quot;:&quot;diffusers.PikaH3FlexModularPipeline&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexBlocks&quot;,&quot;local&quot;:&quot;diffusers.PikaH3FlexBlocks&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexImageReference&quot;,&quot;local&quot;:&quot;diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexVideoReference&quot;,&quot;local&quot;:&quot;diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexAudioReference&quot;,&quot;local&quot;:&quot;diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference&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/225.C7fDjxOl.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/bKdwVYpU.js" rel="modulepreload">
		<link href="/docs/diffusers/main/en/_app/immutable/chunks/D0iXbNom.js" rel="modulepreload">
		<!--13wh65h--><meta name="hf:doc:metadata" content="{&quot;title&quot;:&quot;Pika H3 Flex&quot;,&quot;local&quot;:&quot;pika-h3-flex&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Loading&quot;,&quot;local&quot;:&quot;loading&quot;,&quot;sections&quot;:[{&quot;title&quot;:&quot;Reusing a loaded MiniMax-H3&quot;,&quot;local&quot;:&quot;reusing-a-loaded-minimax-h3&quot;,&quot;sections&quot;:[],&quot;depth&quot;:3}],&quot;depth&quot;:2},{&quot;title&quot;:&quot;What differs from MiniMax-H3&quot;,&quot;local&quot;:&quot;what-differs-from-minimax-h3&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;Generation&quot;,&quot;local&quot;:&quot;generation&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexModularPipeline&quot;,&quot;local&quot;:&quot;diffusers.PikaH3FlexModularPipeline&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexBlocks&quot;,&quot;local&quot;:&quot;diffusers.PikaH3FlexBlocks&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexImageReference&quot;,&quot;local&quot;:&quot;diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexVideoReference&quot;,&quot;local&quot;:&quot;diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference&quot;,&quot;sections&quot;:[],&quot;depth&quot;:2},{&quot;title&quot;:&quot;PikaH3FlexAudioReference&quot;,&quot;local&quot;:&quot;diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference&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="pika-h3-flex" 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="#pika-h3-flex"><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><!---->Pika H3 Flex<!----></span></h1><!--]--><!----> <div class="flex flex-wrap space-x-1"><a href="https://huggingface.co/docs/diffusers/main/en/tutorials/using_peft_for_inference" target="_blank" rel="noopener"><img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/></a></div> <p><a href="https://huggingface.co/Pika-Labs/Pika-H3-Flex" rel="nofollow">Pika H3 Flex</a> は、Pika による <a href="./minimax_h3">MiniMax-H3</a> のポストトレーニングであり、6 ステップの動画・音声生成器です。MiniMax-H3 のパックシーケンス、Qwen3-VL 条件付モデル、および 2 つの VAE を維持したまま、2 つの変更を加えています。1 つ目はサンプラーで、Euler 更新の代わりに、各ステップでクリーンな予測を再ノイズ化する固定の 6 ステップスケジュールを採用しています。2 つ目は重みで、MiniMax-H3 の 2 つのトランスフォーマー分割に対して、ランク 128 の LoRA アダプターをペアで適用しています。動画と同期したステレオサウンドトラックは、同じ 6 回のフォワードパスから出力されます。</p> <p>リリースされるのはアダプターです。<code>fl2va/adapter.safetensors</code> の FL2VA アダプターは、テキストおよびエンドポイントフレームの条件付けのために <code>transformer/</code> パーティションを駆動し、<code>ref2va/adapter.safetensors</code> の Ref2VA アダプターは、マルチモーダル参照のために <code>transformer_ref/</code> を駆動します。ベースコンポーネントは <a href="https://huggingface.co/MiniMaxAI/MiniMax-H3" rel="nofollow">MiniMaxAI/MiniMax-H3</a> のものであり、変更されていません。リポジトリにはそれらのコピーが含まれており、単体で読み込むことができます。</p> <p>Pika H3 Flex は、MiniMax-H3 と同様に、<a href="../../modular_diffusers/overview">Modular Diffusers</a> ブロックのみとして統合されています。つまり、ブロック群とその <a href="/docs/diffusers/main/ja/api/pipelines/pika_h3_flex#diffusers.PikaH3FlexModularPipeline">PikaH3FlexModularPipeline</a> が統合の全体を構成しており、<code>DiffusionPipeline</code> 側の半分は存在しません。</p> <!--[1--><h2 class="relative group"><a id="loading" 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"><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><!---->読み込み中<!----></span></h2><!--]--><!----> <p>パイプラインは2つのステップで組み立てられます。まず、<code>modular_model_index.json</code> を通じて、MiniMax-H3 のベースコンポーネントと2つのスケジューラ設定を読み込みます。次に、選択したワークフローのアダプタを <a href="/docs/diffusers/main/ja/api/loaders/lora#diffusers.loaders.MiniMaxH3LoraLoaderMixin.load_lora_weights">load_lora_weights()</a> で上に読み込みます。これらのアダプタは、diffusers のモジュール名をそのまま使った通常の PEFT ダンプであり、学習時のスケール（<code>alpha == rank</code>、1.0）で読み込まれます。</p> <table><thead><tr><th>ワークフロー</th><th>入力</th><th>Transformerのパーティション</th><th>アダプター</th></tr></thead><tbody><tr><td><code>t2va</code></td><td><code>prompt</code></td><td><code>transformer/</code></td><td><code>fl2va/adapter.safetensors</code></td></tr><tr><td><code>fl2va</code></td><td><code>prompt</code> および <code>image</code> と/または <code>last_image</code></td><td><code>transformer/</code></td><td><code>fl2va/adapter.safetensors</code></td></tr><tr><td><code>ref2va</code></td><td><code>prompt</code> と <code>references</code></td><td><code>transformer_ref/</code></td><td><code>ref2va/adapter.safetensors</code></td></tr></tbody></table> <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> コピーしました</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> ModularPipeline

pipe = ModularPipeline.from_pretrained(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, workflow=<span class="hljs-string">&quot;fl2va&quot;</span>)
pipe.load_components(dtype={<span class="hljs-string">&quot;vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;audio_vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;default&quot;</span>: torch.bfloat16})
pipe.load_lora_weights(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, weight_name=<span class="hljs-string">&quot;fl2va/adapter.safetensors&quot;</span>)<!----></pre></div><!----> <p>workflow を渡すと、ブロックがそのタスク用に剪定されるため、該当する transformer パーティションのみが取得され、アダプターは存在する 1 つのパーティションにルーティングされ、パイプラインはそのタスクのリクエストのみを受け付けます。<code>fl2va</code> パイプラインはキーフレームを必要とし、<code>t2va</code> パイプラインはキーフレームを必要としません。両方のパーティションを保持するパイプライン（<code>workflow=</code> なしで読み込まれたもの）では、Ref2VA アダプターをどこに配置するかを指定する必要があります：<code>pipe.load_lora_weights(..., weight_name="ref2va/adapter.safetensors", load_into_transformer_ref=True)</code>。2 つの VAE は float32 でリリースされており、デコードも float32 で行われます（リリースされたパッケージと同様）。それ以外はすべて bfloat16 で実行されます。</p> <blockquote class="tip"><p><code>pipe.doc</code> は、対象の pipeline が受け取る入力と返す出力を表示します。各入力のデフォルト値、期待されるコンポーネント、および生成される出力がすべて含まれます。</p></blockquote> <!--[2--><h3 class="relative group"><a id="reusing-a-loaded-minimax-h3" 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="#reusing-a-loaded-minimax-h3"><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><!---->読み込んだMiniMax-H3の再利用<!----></span></h3><!--]--><!----> <p>ベースとなるコンポーネントはMiniMax-H3のものです。したがって、すでに読み込まれているMiniMax-H3のパイプラインには、2つのスケジューラとアダプターを除くすべてのコンポーネントが含まれています。このセクションでは、これらの読み込まれたコンポーネントからPika H3 Flexパイプラインを組み立て、両方のパイプラインを並行して利用可能な状態に保ちます。1つの<a href="/docs/diffusers/main/ja/api/modular_diffusers/pipeline_components#diffusers.ComponentsManager">ComponentsManager</a>が各コンポーネントを1回だけ保持し、両方のパイプライン向けにオフロードを行い、Pika H3 Flexパイプラインは2つのスケジューラのみを追加し、アダプターが共有トランスフォーマーがどちらのモデルで実行されるかを決定します。</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> コピーしました</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> ComponentsManager, ModularPipeline

dtype = {<span class="hljs-string">&quot;vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;audio_vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;default&quot;</span>: torch.bfloat16}
manager = ComponentsManager()
manager.enable_auto_cpu_offload(device=<span class="hljs-string">&quot;cuda&quot;</span>)

<span class="hljs-comment"># The MiniMax-H3 pipeline already in use, with every workflow loaded. `collection` registers its components in the</span>
<span class="hljs-comment"># manager under that name, so they can be fetched by it later.</span>
minimax = ModularPipeline.from_pretrained(<span class="hljs-string">&quot;MiniMaxAI/MiniMax-H3&quot;</span>, components_manager=manager, collection=<span class="hljs-string">&quot;minimax-h3&quot;</span>)
minimax.load_components(dtype=dtype)

<span class="hljs-comment"># The Pika H3 Flex pipeline, in its own collection. Only its own components are loaded: the two schedulers.</span>
pika = ModularPipeline.from_pretrained(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, components_manager=manager, collection=<span class="hljs-string">&quot;pika-h3-flex&quot;</span>)
pika.load_components(names=[<span class="hljs-string">&quot;scheduler&quot;</span>, <span class="hljs-string">&quot;audio_scheduler&quot;</span>])

<span class="hljs-comment"># The rest comes from the MiniMax-H3 collection, and `update_components` puts it in the Pika H3 Flex pipeline.</span>
shared = [<span class="hljs-string">&quot;text_encoder&quot;</span>, <span class="hljs-string">&quot;tokenizer&quot;</span>, <span class="hljs-string">&quot;processor&quot;</span>, <span class="hljs-string">&quot;vae&quot;</span>, <span class="hljs-string">&quot;audio_vae&quot;</span>, <span class="hljs-string">&quot;transformer&quot;</span>, <span class="hljs-string">&quot;transformer_ref&quot;</span>]
pika.update_components(**{name: manager.get_one(name=name, collection=<span class="hljs-string">&quot;minimax-h3&quot;</span>) <span class="hljs-keyword">for</span> name <span class="hljs-keyword">in</span> shared})

<span class="hljs-comment"># The adapters go into the shared transformers.</span>
pika.load_lora_weights(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, weight_name=<span class="hljs-string">&quot;fl2va/adapter.safetensors&quot;</span>)
pika.load_lora_weights(
    <span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, weight_name=<span class="hljs-string">&quot;ref2va/adapter.safetensors&quot;</span>, load_into_transformer_ref=<span class="hljs-literal">True</span>
)<!----></pre></div><!----> <p><code>print(manager)</code> を実行すると、結果が表示されます。共有コンポーネントはそれぞれ1回ずつリストに示され、両方のコレクションに属しています。Pika H3 Flex コレクションは2つのスケジューラーを追加し、2つのtransformersにはアダプターが保持されています。</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> コピーしました</div><!----></button><!----></div> <pre class="language-text "><!---->Models:
Name_ID               | Class                           | Size (GB) | Collection
transformer_1397...   | MiniMaxH3Transformer3DModel     | 63.02     | minimax-h3
                      |                                 |           | pika-h3-flex
transformer_ref_13... | MiniMaxH3Transformer3DModel     | 63.02     | minimax-h3
                      |                                 |           | pika-h3-flex
text_encoder_1397...  | Qwen3VLForConditionalGeneration | 62.13     | minimax-h3
                      |                                 |           | pika-h3-flex
vae_1397...           | AutoencoderKLMiniMaxH3          | 9.70      | minimax-h3
                      |                                 |           | pika-h3-flex
audio_vae_1397...     | AutoencoderKLMiniMaxH3Audio     | 0.56      | minimax-h3
                      |                                 |           | pika-h3-flex

Other Components:
ID                        | Class              | Collection
scheduler_1397...         | MiniMaxH3Scheduler | minimax-h3
audio_scheduler_1397...   | MiniMaxH3Scheduler | minimax-h3
scheduler_1397...         | MiniMaxH3Scheduler | pika-h3-flex
audio_scheduler_1397...   | MiniMaxH3Scheduler | pika-h3-flex
tokenizer_1397...         | Qwen2Tokenizer     | minimax-h3
                          |                    | pika-h3-flex
processor_1397...         | Qwen3VLProcessor   | minimax-h3
                          |                    | pika-h3-flex

Additional Component Info:
transformer_1397...:
  Adapters: [&#x27;default_0&#x27;]
transformer_ref_13...:
  Adapters: [&#x27;default_0&#x27;]<!----></pre></div><!----> <p>アダプターが、共有されている transformers がどのモデルであるかを決定します。MiniMax-H3 に戻すには、<code>pika.disable_lora()</code> を呼び出すだけです。次の <code>minimax(...)</code> リクエストではベースモデルが実行されます。<code>pika.enable_lora()</code> を呼び出すと、Pika H3 Flex が再度有効になります。その間に何も再読み込みされることはなく、いずれかのパイプラインが各コンポーネントに到達するたびに、マネージャーがそれらをアクセラレーターにオン/オフします。</p> <!--[1--><h2 class="relative group"><a id="what-differs-from-minimax-h3" 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="#what-differs-from-minimax-h3"><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><!---->MiniMax-H3との違い<!----></span></h2><!--]--><!----> <ul><li><strong>ステップ数は6、<code>num_inference_steps</code>は不要です。</strong> <a href="/docs/diffusers/main/ja/api/schedulers/minimax_h3#diffusers.MiniMaxH3Scheduler">MiniMaxH3Scheduler</a>の両インスタンス（動画用は<code>shift=12.0</code>、音声用は<code>shift=3.0</code>）は、パイプラインの<code>denoising_step_list</code>設定、すなわち公開されているスケジュール<code>[1000, 750, 500, 250, 100, 40, 0]</code>に対して<code>stochastic_sampling</code>で実行されます。各ステップでクリーンなサンプルを予測し、最終ステップを除くすべてのステップでは、次のsigmaまで新しいノイズで再ノイズ化します。ステップ数はアダプタの契約事項であり、入力パラメータではありません。</li> <li><strong>各ステップで新しいノイズを生成します。</strong> リクエストは<code>generator</code>から固定の順序でサンプリングします。条件付けノイズにはキーフレームまたは視覚参照ごとに1回、次に動画ノイズ、音声ノイズ、そして最初の5ステップそれぞれの動画および音声の再ノイズ化用サンプリングが行われます。<code>latents</code>または<code>audio_latents</code>を渡すと、最初の2回のサンプリングが置き換えられ、残りはそのまま維持されます。</li> <li><strong>キーフレームは引き伸ばされず、クロップされます。</strong> <code>fl2va</code>のキーフレームは、自身の解像度でキャンバスのアスペクト比に合わせて中央クロップされ、キャンバスにリサイズされます。</li> <li><strong>参照はキャンバスの解像度ティアに従います。</strong> Pika H3 Flex はバケットテーブル上で生成を行います。480pティアと768pティアがあり、それぞれ11種類のアスペクト比に対応しています。<code>ref2va</code> リクエストは、そのキャンバスのティアに基づいて条件付けを行います。画像参照は、768ティア未満では短辺1024ピクセル、それ以上では2048ピクセルでエンコードされ、動画参照は該当ティアのキャンバールールに従います。すべての動画または単体オーディオの参照は2秒から15秒の長さで、動画の再生時間と単体オーディオの再生時間はそれぞれ合計15秒以下である必要があります。</li></ul> <!--[1--><h2 class="relative group"><a id="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="#generation"><span><svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" role="img" width="1em" height="1em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 256 256"><path d="M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z" fill="currentColor"></path></svg><!----></span></a> <span><!---->生成<!----></span></h2><!--]--><!----> <p>リリースされたバケットは、Pika H3 Flex が検証されたキャンバスです。<code>height</code> と <code>width</code> として1つを指定してください。</p> <table><thead><tr><th></th><th>21:9</th><th>16:9</th><th>3:2</th><th>4:3</th><th>5:4</th><th>1:1</th><th>4:5</th><th>3:4</th><th>2:3</th><th>9:16</th><th>9:21</th></tr></thead><tbody><tr><td>480p</td><td>960x416</td><td>832x480</td><td>704x480</td><td>640x480</td><td>608x480</td><td>480x480</td><td>480x608</td><td>480x640</td><td>480x704</td><td>480x832</td><td>416x960</td></tr><tr><td>768p</td><td>1536x672</td><td>1344x768</td><td>1152x768</td><td>1024x768</td><td>960x768</td><td>768x768</td><td>768x960</td><td>768x1024</td><td>768x1152</td><td>768x1344</td><td>672x1536</td></tr></tbody></table> <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> コピーしました</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> ComponentsManager, ModularPipeline
<span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> load_image
<span class="hljs-keyword">from</span> diffusers.utils.export_utils <span class="hljs-keyword">import</span> encode_video

<span class="hljs-comment"># 61.7GB of transformer and 62.1GB of conditioner do not sit on one accelerator, so the components are</span>
<span class="hljs-comment"># registered in a manager that moves each one on and off as the blocks reach it.</span>
manager = ComponentsManager()
manager.enable_auto_cpu_offload(device=<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

<span class="hljs-comment"># Create one pipeline for all supported workflows, then load the `fl2va` workflow&#x27;s components, which is</span>
<span class="hljs-comment"># enough to run both the fl2va and t2va requests below.</span>
pipe = ModularPipeline.from_pretrained(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, components_manager=manager)
pipe.load_components(
    workflow=<span class="hljs-string">&quot;fl2va&quot;</span>, dtype={<span class="hljs-string">&quot;vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;audio_vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;default&quot;</span>: torch.bfloat16}
)
pipe.load_lora_weights(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, weight_name=<span class="hljs-string">&quot;fl2va/adapter.safetensors&quot;</span>)

prompt = (
    <span class="hljs-string">&quot;integrated_multimodal_description: A cinematic close shot of a red ceramic cup on a wooden table beside an &quot;</span>
    <span class="hljs-string">&quot;open window. A soft breeze moves a linen curtain. The camera slowly pushes toward the cup as warm morning &quot;</span>
    <span class="hljs-string">&quot;light crosses the table.\n&quot;</span>
    <span class="hljs-string">&quot;overall_soundscape: Gentle wind through the window and faint birdsong outside. No speech.\n&quot;</span>
    <span class="hljs-string">&quot;non_diegetic_music: None.\n&quot;</span>
)
outputs = [<span class="hljs-string">&quot;videos&quot;</span>, <span class="hljs-string">&quot;audio&quot;</span>, <span class="hljs-string">&quot;sampling_rate&quot;</span>]

<span class="hljs-comment"># Text to video + audio on the 480p 16:9 bucket, 5 seconds.</span>
results = pipe(
    prompt=prompt, 
    height=<span class="hljs-number">480</span>, 
    width=<span class="hljs-number">832</span>, 
    num_frames=<span class="hljs-number">124</span>, 
    generator=torch.Generator(<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">42</span>), 
    output=outputs
)

<span class="hljs-comment"># First frame (and optionally `last_image`) to video + audio. The keyframe is center-cropped onto the canvas and</span>
<span class="hljs-comment"># held as the first frame, so the prompt describes the keyframe and what happens from it.</span>
image = load_image(
    <span class="hljs-string">&quot;https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/astronaut.jpg&quot;</span>
)
i2va_prompt = (
    <span class="hljs-string">&quot;For the target video, at 0.00 seconds into the target video, &lt;Picture 1&gt; (from [Shot 1]) is fully referenced.\n\n&quot;</span>
    <span class="hljs-string">&quot;integrated_multimodal_description: [Shot 1] Begin with the supplied image of a small toy astronaut figurine &quot;</span>
    <span class="hljs-string">&quot;standing on a grey rocky moon surface, holding a white egg-shaped object. Preserve the figurine, its orange &quot;</span>
    <span class="hljs-string">&quot;visor and the rocks. The astronaut slowly turns its helmet toward the camera and lifts the egg a little; fine &quot;</span>
    <span class="hljs-string">&quot;dust drifts up around its boots. The camera slowly pushes in in one continuous shot. By the end the astronaut &quot;</span>
    <span class="hljs-string">&quot;stands still, egg held up.\n\n&quot;</span>
    <span class="hljs-string">&quot;overall_soundscape: Soft low wind and faint crunch of dust under the boots. No speech.\n\n&quot;</span>
    <span class="hljs-string">&quot;non_diegetic_music: N/A\n&quot;</span>
)
results = pipe(
    prompt=i2va_prompt, 
    image=image, 
    height=<span class="hljs-number">480</span>, 
    width=<span class="hljs-number">832</span>, 
    num_frames=<span class="hljs-number">124</span>,
    generator=torch.Generator(<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">42</span>), 
    output=outputs,
)

encode_video(
    results[<span class="hljs-string">&quot;videos&quot;</span>][<span class="hljs-number">0</span>],
    fps=<span class="hljs-number">24</span>,
    output_path=<span class="hljs-string">&quot;pika_h3_flex_fl2va.mp4&quot;</span>,
    audio=results[<span class="hljs-string">&quot;audio&quot;</span>][<span class="hljs-number">0</span>],
    audio_sample_rate=results[<span class="hljs-string">&quot;sampling_rate&quot;</span>],
)<!----></pre></div><!----> <p><code>ref2va</code>ワークフローは、MiniMax-H3と同様に、最大9枚の画像、3本の動画、3本の音声クリップ（合計12件）からなる順序付きの参照リストを受け取ります。モダリティごとに1つの参照クラスが割り当てられます。また、このリクエストはメモリ消費量が最も大きくなります。参照はシーケンスに独自の行を追加するため、768pではtransformer単体で80 GBのカードを埋め尽くします。そのため、以下の例では、transformerのブロックをCPUメモリから1つずつストリーミングし、マネージャーは従来どおり他のコンポーネントをデバイスに移動・退避します。この設定では、1枚の80 GBカードで768pの参照リクエストを実行できます。ストリーミングを行わない場合は、480pに制限してください。</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> コピーしました</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> ComponentsManager, ModularPipeline
<span class="hljs-keyword">from</span> diffusers.modular_pipelines.pika_h3_flex <span class="hljs-keyword">import</span> (
    PikaH3FlexAudioReference,
    PikaH3FlexImageReference,
    PikaH3FlexVideoReference,
)

manager = ComponentsManager()
manager.enable_auto_cpu_offload(device=<span class="hljs-string">&quot;cuda&quot;</span>)  <span class="hljs-comment"># or &quot;mps&quot;, &quot;xpu&quot;, &quot;cpu&quot;</span>

pipe = ModularPipeline.from_pretrained(
    <span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, workflow=<span class="hljs-string">&quot;ref2va&quot;</span>, components_manager=manager
)
pipe.load_components(dtype={<span class="hljs-string">&quot;vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;audio_vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;default&quot;</span>: torch.bfloat16})
pipe.load_lora_weights(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, weight_name=<span class="hljs-string">&quot;ref2va/adapter.safetensors&quot;</span>)

<span class="hljs-comment"># Keep the transformer in CPU memory and bring in one block at a time while it runs.</span>
pipe.transformer_ref.enable_group_offload(
    onload_device=torch.device(<span class="hljs-string">&quot;cuda&quot;</span>),
    offload_device=torch.device(<span class="hljs-string">&quot;cpu&quot;</span>),
    offload_type=<span class="hljs-string">&quot;block_level&quot;</span>,
    num_blocks_per_group=<span class="hljs-number">1</span>,
    use_stream=<span class="hljs-literal">True</span>,
)

results = pipe(
    prompt=<span class="hljs-string">&quot;subject_definitions: &lt;Subject 1&gt; is the woman in &lt;Picture 1&gt;. summary: [reference generation] &lt;Subject 1&gt; waves at the camera.&quot;</span>,
    references=[
        PikaH3FlexImageReference.from_file(<span class="hljs-string">&quot;subject.png&quot;</span>),
        PikaH3FlexVideoReference.from_file(<span class="hljs-string">&quot;motion_ref.mp4&quot;</span>),  <span class="hljs-comment"># frames, their `fps` and the soundtrack</span>
        PikaH3FlexAudioReference.from_file(<span class="hljs-string">&quot;voice.wav&quot;</span>),  <span class="hljs-comment"># waveform and its `sample_rate`</span>
    ],
    height=<span class="hljs-number">768</span>,
    width=<span class="hljs-number">1344</span>,
    num_frames=<span class="hljs-number">124</span>,
    generator=torch.Generator(<span class="hljs-string">&quot;cuda&quot;</span>).manual_seed(<span class="hljs-number">42</span>),
    output=[<span class="hljs-string">&quot;videos&quot;</span>, <span class="hljs-string">&quot;audio&quot;</span>, <span class="hljs-string">&quot;sampling_rate&quot;</span>],
)<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="diffusers.PikaH3FlexModularPipeline" 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="#diffusers.PikaH3FlexModularPipeline"><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><!---->PikaH3FlexModularPipeline<!----></span></h2><!--]--><!----> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.PikaH3FlexModularPipeline"><!----><h3 class="!m-0"><span class="flex-1 break-all md:text-lg bg-gradient-to-r px-2.5 py-1.5 rounded-xl from-indigo-50/70 to-white dark:from-gray-900 dark:to-gray-950 dark:text-indigo-300 text-indigo-700"><svg class="mr-1.5 text-indigo-500 inline-block -mt-0.5" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width=".8em" height=".8em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 24 24"><path class="uim-quaternary" d="M20.23 7.24L12 12L3.77 7.24a1.98 1.98 0 0 1 .7-.71L11 2.76c.62-.35 1.38-.35 2 0l6.53 3.77c.29.173.531.418.7.71z" opacity=".25" fill="currentColor"></path><path class="uim-tertiary" d="M12 12v9.5a2.09 2.09 0 0 1-.91-.21L4.5 17.48a2.003 2.003 0 0 1-1-1.73v-7.5a2.06 2.06 0 0 1 .27-1.01L12 12z" opacity=".5" fill="currentColor"></path><path class="uim-primary" d="M20.5 8.25v7.5a2.003 2.003 0 0 1-1 1.73l-6.62 3.82c-.275.13-.576.198-.88.2V12l8.23-4.76c.175.308.268.656.27 1.01z" fill="currentColor"></path></svg><span class="font-light">class</span> <span class="font-medium">diffusers.</span><span class="font-semibold">PikaH3FlexModularPipeline</span></span></h3><!----> <a id="diffusers.PikaH3FlexModularPipeline" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.PikaH3FlexModularPipeline"><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><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/modular_pipelines/pika_h3_flex/modular_pipeline.py#L211" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">blocks<span class="opacity-60">: diffusers.modular_pipelines.modular_pipeline.ModularPipelineBlocks | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">pretrained_model_name_or_path<span class="opacity-60">: str | os.PathLike | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">components_manager<span class="opacity-60">: diffusers.modular_pipelines.components_manager.ComponentsManager | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">collection<span class="opacity-60">: str | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">workflow<span class="opacity-60">: str | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">modular_config_dict<span class="opacity-60">: dict[str, typing.Any] | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">config_dict<span class="opacity-60">: dict[str, typing.Any] | None = None</span></span></span><span class="comma cursor-default"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">**kwargs<span class="opacity-60"></span></span></span><!--]--> <span>)</span> <!--[-1--><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>Pika H3 Flex と、ポストトレーニング済みの6ステップ MiniMax-H3 を用いた、動画と音声の同時生成を行う ModularPipeline です。<code>transformer/</code> チェックポイントのパーティションに対しては <code>fl2va</code> アダプタを、<code>transformer_ref/</code> に対しては <code>ref2va</code> アダプタを使用し、<code>t2va</code>（テキストのみ）および <code>fl2va</code>（先頭および/または末尾キーフレーム）ワークフロー、ならびに <code>ref2va</code>（オムニリファレンス）ワークフローに対応しています。</p> <p>ベースコンポーネントは MiniMax-H3 のものを使用し、Pika H3 Flex のリリースは、2 つの transformer パーティション上に構築された rank-128 の LoRA アダプターの一対です。そのため、パイプラインは 2 ステップで組み立てられます。まず、ベースコンポーネントと 2 つのスケジューラを指定する <code>modular_model_index.json</code> を通じてコンポーネントを読み込み、次に、選択したワークフローのアダプターを <a href="/docs/diffusers/main/ja/api/loaders/lora#diffusers.loaders.MiniMaxH3LoraLoaderMixin.load_lora_weights">load_lora_weights()</a> で読み込みます。</p> <div class="relative group rounded-md"><a id="diffusers.PikaH3FlexModularPipeline.example" class="header-link block pr-0.5 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="#diffusers.PikaH3FlexModularPipeline.example"><span><svg class="text-smd" 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> <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> コピーしました</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> ModularPipeline

pipe = ModularPipeline.from_pretrained(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, workflow=<span class="hljs-string">&quot;fl2va&quot;</span>)
pipe.load_components(dtype={<span class="hljs-string">&quot;vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;audio_vae&quot;</span>: torch.float32, <span class="hljs-string">&quot;default&quot;</span>: torch.bfloat16})
pipe.load_lora_weights(<span class="hljs-string">&quot;Pika-Labs/Pika-H3-Flex&quot;</span>, weight_name=<span class="hljs-string">&quot;fl2va/adapter.safetensors&quot;</span>)<!----></pre></div><!----></div><!----> <p>MiniMax-H3と同様に、Pika H3 Flexは、テキストの条件付け、キーフレームまたは参照の条件付け潜在変数、音声潜在変数、動画潜在変数を一度に保持する<strong>1つのパックされたシーケンス</strong>をデノイズします。そのため、ブロックはモダリティごとのテンソルではなく、行レイアウトを渡り回し、パイプラインは2つのスケジューラ（動画用は<code>shift = 12.0</code>、音声用は<code>shift = 3.0</code>）を持ち、これらは単一のtransformer呼び出しの中でステップを進めます。MiniMax-H3と異なり、この2つのスケジューラは、固定の6ステップの<code>denoising_step_list</code>に対して<code>stochastic_sampling</code>を行う<a href="/docs/diffusers/main/ja/api/schedulers/minimax_h3#diffusers.MiniMaxH3Scheduler">MiniMaxH3Scheduler</a>のインスタンスです。<code>num_inference_steps</code>は存在せず、各ステップではリクエストのジェネレーターから新しいノイズがサンプリングされます。</p> <p>アダプターはベースモデルと同様にガイダンス蒸留されており、<code>negative_prompt</code>も<code>guidance_scale</code>も存在せず、各ステップでフォワードパスが正確に1回だけ実行されます。</p> <p>このモジュールは、各ブロックが参照するモデルの事実を保持します。config から導出される幾何学的情報はプロパティとして、キャンバス、ティア、およびフレーム数の算出は関数として定義されています。また、条件付け、エンコーディング、ノイズの契約は、ブロック自体に実装されています。</p></div> <!--[1--><h2 class="relative group"><a id="diffusers.PikaH3FlexBlocks" 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="#diffusers.PikaH3FlexBlocks"><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><!---->PikaH3FlexBlocks<!----></span></h2><!--]--><!----> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.PikaH3FlexBlocks"><!----><h3 class="!m-0"><span class="flex-1 break-all md:text-lg bg-gradient-to-r px-2.5 py-1.5 rounded-xl from-indigo-50/70 to-white dark:from-gray-900 dark:to-gray-950 dark:text-indigo-300 text-indigo-700"><svg class="mr-1.5 text-indigo-500 inline-block -mt-0.5" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width=".8em" height=".8em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 24 24"><path class="uim-quaternary" d="M20.23 7.24L12 12L3.77 7.24a1.98 1.98 0 0 1 .7-.71L11 2.76c.62-.35 1.38-.35 2 0l6.53 3.77c.29.173.531.418.7.71z" opacity=".25" fill="currentColor"></path><path class="uim-tertiary" d="M12 12v9.5a2.09 2.09 0 0 1-.91-.21L4.5 17.48a2.003 2.003 0 0 1-1-1.73v-7.5a2.06 2.06 0 0 1 .27-1.01L12 12z" opacity=".5" fill="currentColor"></path><path class="uim-primary" d="M20.5 8.25v7.5a2.003 2.003 0 0 1-1 1.73l-6.62 3.82c-.275.13-.576.198-.88.2V12l8.23-4.76c.175.308.268.656.27 1.01z" fill="currentColor"></path></svg><span class="font-light">class</span> <span class="font-medium">diffusers.</span><span class="font-semibold">PikaH3FlexBlocks</span></span></h3><!----> <a id="diffusers.PikaH3FlexBlocks" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.PikaH3FlexBlocks"><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><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/modular_pipelines/pika_h3_flex/modular_blocks_pika_h3_flex.py#L668" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><!--]--> <span>)</span> <!--[-1--><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>Pika H3 Flex による動画と音声の同時生成に対応した Auto Modular パイプラインブロックです。<code>t2va</code>（テキストのみ）、<code>fl2va</code>（先頭および/または末尾キーフレーム）、<code>ref2va</code>（オムニリファレンス）のワークフローがあり、<code>references</code> とキーフレームの入力によって選択されます。<code>t2va</code> と <code>fl2va</code> ワークフローは <code>transformer</code> パーティションを <code>fl2va</code> アダプターで実行し、<code>ref2va</code> は <code>transformer_ref</code> パーティションを <code>ref2va</code> アダプターで実行します。そのため、リクエストが採用するワークフローのアダプターを読み込む必要があります。<code>workflow=</code> を指定しない場合、コンポーネントの読み込み時に 61.7GB の <strong>両方の</strong> transformer パーティションが読み込まれてしまいます。タスクに必要なもののみを読み込むには、一方を指定してください。</p> <p>サポートされているワークフロー:</p> <ul><li><code>t2va</code>: <code>prompt</code>が必要です</li> <li><code>fl2va</code>: <code>prompt</code>、<code>image</code>、または <code>prompt</code>、<code>last_image</code> が必要です</li> <li><code>ref2va</code>: <code>prompt</code>、<code>references</code> が必要です</li></ul> <p>コンポーネント:
image_processor (<code>VaeImageProcessor</code>) text_encoder (<code>Qwen3VLForConditionalGeneration</code>) tokenizer
(<code>Qwen2Tokenizer</code>) processor (<code>Qwen3VLProcessor</code>) vae (<code>AutoencoderKLMiniMaxH3</code>) audio_vae
(<code>AutoencoderKLMiniMaxH3Audio</code>) scheduler (<code>MiniMaxH3Scheduler</code>) audio_scheduler (<code>MiniMaxH3Scheduler</code>)
transformer_ref (<code>MiniMaxH3Transformer3DModel</code>) transformer (<code>MiniMaxH3Transformer3DModel</code>) video_processor
(<code>VideoProcessor</code>)</p> <p>設定:
 (デフォルト: 768)  (デフォルト: 1032192)  (デフォルト: [1000, 750, 500, 250, 100, 40, 0])</p> <p>Inputs:
references (<code>list</code>, <em>任意</em>):
条件付けに使用する参照です。<strong>モデルがそれらを参照する順序</strong>で指定します。順序はプロンプト提示におけるラベル付けと、共有ロータリークロック上での配置を決定するため、順序が異なれば異なるリクエストとなります。モダリティごとに1つのデータクラスを持ち、すべてメモリ内のメディアを保持します。<code>PikaH3FlexImageReference</code>（最大9個）、独自の<code>fps</code>を持つ<code>PikaH3FlexVideoReference</code>（最大3個、その<code>audio</code>サウンドトラックも条件付けの対象）、または独自の<code>sample_rate</code>を持つ<code>PikaH3FlexAudioReference</code>（最大3個）です。参照の合計は最大12個までで、オーディオ参照のみでは構成できません。すべての動画または単独のオーディオ参照は2秒から15秒の長さで、動画の再生時間と単独オーディオの再生時間はそれぞれ合計15秒以下である必要があります。これらのブロックはメディアファイルを開きません。各クラスの<code>from_file</code>クラスメソッドでデコードし、レートも一緒に読み込みます。
height (<code>int</code>, <em>任意</em>):
生成される動画のピクセル単位の縦幅です。32の倍数である必要があります。
width (<code>int</code>, <em>任意</em>):
生成される動画のピクセル単位の横幅です。32の倍数である必要があります。
num_frames (<code>int</code>, <em>任意</em>):
固定の24 fpsで生成するフレーム数です。120から360、つまり5秒から15秒の範囲で、動画VAEがデコードできる次の<code>17 * n + 5</code>に切り上げられます。参照サウンドトラックと同じ長さの動画を生成するには、<code>round(samples / sample_rate * 24)</code>を渡します。
image (<code>Image</code>, <em>任意</em>):
動画の開始となるキーフレームです。キャンバスのアスペクト比に合わせて中央クロップされ、それにリサイズされます。キャンバスはデフォルトで、そのキーフレーム自身のアスペクト比から導出されます。
last_image (<code>Image</code>, <em>任意</em>):
動画の終了となるキーフレームです。<em>最大</em>のフレームまで生成するために単独で渡すこともできます。<code>image</code>と同様に、キャンバスのアスペクト比に合わせて中央クロップされ、それにリサイズされます。
prompt (<code>str</code>):
生成をガイドするプロンプトです。単一の文字列です。
normalized_references (<code>list</code>, <em>任意</em>):
セットアップステップで正規化された参照です。パックされた順序で保持されます。
keyframes (<code>list</code>, <em>任意</em>):
ターゲットキャンバス上に配置されたキーフレームです。パックされた順序で保持されます。
condition_latents (<code>list</code>, <em>任意</em>):
エンコードされた動画条件付けの潜変数です。パックされた順序で、画像参照と動画参照ごとに1つです。その形状は、すべての参照ブロックのジオメトリが由来するものです。
audio_condition_latents (<code>list</code>, <em>任意</em>):
エンコードされたオーディオ条件付けの行です。パックされた順序で、オーディオを持つ参照ごとに1つです。
generator (<code>Generator</code>, <em>任意</em>):
リクエストのジェネレーターです。条件付けノイズは、生成される行のノイズよりも先に、条件ごとに1回ずつ、このジェネレーターから抽出されます。
latents (<code>Tensor</code>):
形状<code>(1, 24, num_latent_frames, latent_height, latent_width)</code>の事前生成された動画ノイズです。抽出の代わりに使用されます。
audio_latents (<code>Tensor</code>):
形状<code>(2, 32, num_audio_latents)</code>の事前生成されたオーディオノイズです。
*<em>denoiser_input_fields (<code>None</code>,</em>optional<em>):
トランスフォーマーが名前によって読み取るパックされたシーケンスの構造記述です：<code>token_tags</code>、<code>position_ids</code>、および3つの行インデックステンソル。
attention_kwargs (<code>dict</code>,</em>optional<em>):
アテンションプロセッサ用の追加のkwargsです。
keyframe_anchors (<code>tuple</code>,</em>optional<em>, デフォルトは ()):
パックされた順序で、各キーフレームが動画のどの端にアンカーされるかを指定します。
output_type (<code>str</code>,</em>optional*, デフォルトは pil):
出力形式です。‘pil’、‘np’、または‘pt’です。</p> <p>出力:
videos (<code>list</code>):
生成された動画。
audio (<code>Tensor</code>):
生成されたサウンドトラック。形状は <code>(1, 2, num_samples)</code> です。
sampling_rate (<code>int</code>):
生成されたサウンドトラックのサンプリングレート (Hz)。</p></div> <!--[1--><h2 class="relative group"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference" 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference"><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><!---->PikaH3FlexImageReference<!----></span></h2><!--]--><!----> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference"><!----><h3 class="!m-0"><span class="flex-1 break-all md:text-lg bg-gradient-to-r px-2.5 py-1.5 rounded-xl from-indigo-50/70 to-white dark:from-gray-900 dark:to-gray-950 dark:text-indigo-300 text-indigo-700"><svg class="mr-1.5 text-indigo-500 inline-block -mt-0.5" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width=".8em" height=".8em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 24 24"><path class="uim-quaternary" d="M20.23 7.24L12 12L3.77 7.24a1.98 1.98 0 0 1 .7-.71L11 2.76c.62-.35 1.38-.35 2 0l6.53 3.77c.29.173.531.418.7.71z" opacity=".25" fill="currentColor"></path><path class="uim-tertiary" d="M12 12v9.5a2.09 2.09 0 0 1-.91-.21L4.5 17.48a2.003 2.003 0 0 1-1-1.73v-7.5a2.06 2.06 0 0 1 .27-1.01L12 12z" opacity=".5" fill="currentColor"></path><path class="uim-primary" d="M20.5 8.25v7.5a2.003 2.003 0 0 1-1 1.73l-6.62 3.82c-.275.13-.576.198-.88.2V12l8.23-4.76c.175.308.268.656.27 1.01z" fill="currentColor"></path></svg><span class="font-light">class</span> <span class="font-medium">diffusers.modular_pipelines.pika_h3_flex.</span><span class="font-semibold">PikaH3FlexImageReference</span></span></h3><!----> <a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference"><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><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/modular_pipelines/pika_h3_flex/references.py#L83" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">image<span class="opacity-60">: typing.Union[PIL.Image.Image, numpy.ndarray, torch.Tensor]</span></span></span><!--]--> <span>)</span> <!--[-1--><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[0--><p class="flex items-center font-semibold !mt-2 !mb-2 text-gray-800">パラメータ <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700 ml-3"></span></p> <ul class="px-2"><!--[--><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference.image" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference.image"><span><svg class="text-smd" 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><!----><strong>image</strong> (<code>PIL.Image.Image</code>、<code>np.ndarray</code> または <code>torch.Tensor</code>) —
参照画像です。画像、<code>(height, width, 3)</code> の配列、または <code>(3, height, width)</code> のテンソルで、<code>uint8</code> または <code>[0, 1]</code> 範囲の浮動小数点数です。生成されるジオメトリには決して束縛されません。これは、リクエストが生成するキャンバスサイズに関わらず、自身の短い辺でエンコードされます（公開されているチェックポイントでは 2048 です）。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>被写体、スタイル、またはシーン参照: 1リクエストあたり最大9件。</p> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference.from_file"><!----><h4 class="!m-0"><span class="flex-1 rounded-xl py-0.5 break-all bg-gradient-to-r from-blue-50/60 to-white dark:from-gray-900 dark:to-gray-950 text-blue-700 dark:text-blue-300 font-medium px-2"><svg width="1em" height="1em" viewBox="0 0 32 33" class="mr-1 inline-block -mt-0.5"  xmlns="http://www.w3.org/2000/svg"><path d="M5.80566 18.3545C4.90766 17.4565 4.90766 16.0005 5.80566 15.1025L14.3768 6.53142C15.2748 5.63342 16.7307 5.63342 17.6287 6.53142L26.1999 15.1025C27.0979 16.0005 27.0979 17.4565 26.1999 18.3545L17.6287 26.9256C16.7307 27.8236 15.2748 27.8236 14.3768 26.9256L5.80566 18.3545Z" fill="currentColor" fill-opacity="0.25"/><path fill-rule="evenodd" clip-rule="evenodd" d="M16.4801 13.9619C16.4801 12.9761 16.7467 12.5436 16.9443 12.3296C17.1764 12.078 17.5731 11.8517 18.2275 11.707C18.8821 11.5623 19.638 11.5342 20.4038 11.5582C20.7804 11.57 21.1341 11.5932 21.4719 11.6156L21.5263 11.6193C21.8195 11.6389 22.1626 11.6618 22.4429 11.6618V7.40825C22.3209 7.40825 22.1219 7.39596 21.7544 7.37149C21.4202 7.34925 20.9976 7.32115 20.5371 7.30672C19.6286 7.27824 18.4672 7.29779 17.3093 7.55377C16.1512 7.8098 14.8404 8.33724 13.8181 9.4452C12.7612 10.5907 12.2266 12.1236 12.2266 13.9619V15.0127H10.6836V19.2662H12.2266V26.6332H16.4801V19.2662H20.3394V15.0127H16.4801V13.9619Z" fill="currentColor"/></svg>from_file</span></h4><!----> <a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference.from_file" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference.from_file"><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><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/modular_pipelines/pika_h3_flex/references.py#L99" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">media<span class="opacity-60"></span></span></span><!--]--> <span>)</span> <!--[-1--><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[0--><p class="flex items-center font-semibold !mt-2 !mb-2 text-gray-800">パラメータ <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700 ml-3"></span></p> <ul class="px-2"><!--[--><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference.from_file.media" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexImageReference.from_file.media"><span><svg class="text-smd" 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><!----><strong>media</strong> (<code>str</code> または <code>os.PathLike</code>) — 画像のパスまたは URL。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p><code>PikaH3FlexImageReference</code>に、<a href="/docs/diffusers/main/ja/api/utilities#diffusers.utils.load_image">load_image()</a>を通じて画像ファイルを読み込みます。</p></div></div> <!--[1--><h2 class="relative group"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference" 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference"><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><!---->PikaH3FlexVideoReference<!----></span></h2><!--]--><!----> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference"><!----><h3 class="!m-0"><span class="flex-1 break-all md:text-lg bg-gradient-to-r px-2.5 py-1.5 rounded-xl from-indigo-50/70 to-white dark:from-gray-900 dark:to-gray-950 dark:text-indigo-300 text-indigo-700"><svg class="mr-1.5 text-indigo-500 inline-block -mt-0.5" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width=".8em" height=".8em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 24 24"><path class="uim-quaternary" d="M20.23 7.24L12 12L3.77 7.24a1.98 1.98 0 0 1 .7-.71L11 2.76c.62-.35 1.38-.35 2 0l6.53 3.77c.29.173.531.418.7.71z" opacity=".25" fill="currentColor"></path><path class="uim-tertiary" d="M12 12v9.5a2.09 2.09 0 0 1-.91-.21L4.5 17.48a2.003 2.003 0 0 1-1-1.73v-7.5a2.06 2.06 0 0 1 .27-1.01L12 12z" opacity=".5" fill="currentColor"></path><path class="uim-primary" d="M20.5 8.25v7.5a2.003 2.003 0 0 1-1 1.73l-6.62 3.82c-.275.13-.576.198-.88.2V12l8.23-4.76c.175.308.268.656.27 1.01z" fill="currentColor"></path></svg><span class="font-light">class</span> <span class="font-medium">diffusers.modular_pipelines.pika_h3_flex.</span><span class="font-semibold">PikaH3FlexVideoReference</span></span></h3><!----> <a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference"><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><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/modular_pipelines/pika_h3_flex/references.py#L111" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">frames<span class="opacity-60">: typing.Union[list[PIL.Image.Image], numpy.ndarray, torch.Tensor]</span></span></span><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">fps<span class="opacity-60">: float | None = None</span></span></span><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">audio<span class="opacity-60">: typing.Optional[torch.Tensor] = None</span></span></span><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">sample_rate<span class="opacity-60">: int | None = None</span></span></span><!--]--> <span>)</span> <!--[-1--><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[0--><p class="flex items-center font-semibold !mt-2 !mb-2 text-gray-800">パラメータ <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700 ml-3"></span></p> <ul class="px-2"><!--[--><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.frames" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.frames"><span><svg class="text-smd" 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><!----><strong>frames</strong> (<code>list[PIL.Image.Image]</code>、<code>np.ndarray</code> または <code>torch.Tensor</code>) —
参照フレーム: 画像のリスト、<code>(num_frames, height, width, 3)</code> の配列、または <code>(num_frames, 3, height, width)</code> のテンソルで、<code>uint8</code> または <code>[0, 1]</code> 範囲の浮動小数点数です。<!----></span></span></li><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.fps" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.fps"><span><svg class="text-smd" 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><!----><strong>fps</strong> (<code>float</code>、<em>任意</em>、デフォルト値は 24.0) —
<code>frames</code> が持つフレームレートです。この値により、参照のビジョンブロックがコンディショナーの 2 fps グリッド上に配置されます。MiniMax-H3 自体のクロックは 24 fps であるため、それ以外のレートは、フレーム全体を削除または複製することで 24 fps にリサンプリングされます。したがって、フレームがファイルから取得された場合は、このフィールドを正しく設定することが重要です。<!----></span></span></li><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.audio" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.audio"><span><svg class="text-smd" 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><!----><strong>audio</strong> (<code>torch.Tensor</code>、形状 <code>(channels, num_samples)</code>、<em>任意</em>) —
この動画のサウンドトラック（モノラルまたはステレオ）です。参照そのものとして、参照の参照としてではなく、条件付けされます。省略した場合、参照はモーションのみを条件として扱います。<!----></span></span></li><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.sample_rate" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.sample_rate"><span><svg class="text-smd" 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><!----><strong>sample_rate</strong> (<code>int</code>, <em>任意</em>) —
<code>audio</code> がサンプルを保持するレートです。省略した場合、オーディオ VAE 自身のレートが使用され、サンプルは変更されません。それ以外のレートが指定された場合は、そのレートにリサンプリングされます。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>モーションとカメラのリファレンス: リクエストごとに最大3つまで、自身のサウンドトラックと合わせて条件付けられます。</p> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.from_file"><!----><h4 class="!m-0"><span class="flex-1 rounded-xl py-0.5 break-all bg-gradient-to-r from-blue-50/60 to-white dark:from-gray-900 dark:to-gray-950 text-blue-700 dark:text-blue-300 font-medium px-2"><svg width="1em" height="1em" viewBox="0 0 32 33" class="mr-1 inline-block -mt-0.5"  xmlns="http://www.w3.org/2000/svg"><path d="M5.80566 18.3545C4.90766 17.4565 4.90766 16.0005 5.80566 15.1025L14.3768 6.53142C15.2748 5.63342 16.7307 5.63342 17.6287 6.53142L26.1999 15.1025C27.0979 16.0005 27.0979 17.4565 26.1999 18.3545L17.6287 26.9256C16.7307 27.8236 15.2748 27.8236 14.3768 26.9256L5.80566 18.3545Z" fill="currentColor" fill-opacity="0.25"/><path fill-rule="evenodd" clip-rule="evenodd" d="M16.4801 13.9619C16.4801 12.9761 16.7467 12.5436 16.9443 12.3296C17.1764 12.078 17.5731 11.8517 18.2275 11.707C18.8821 11.5623 19.638 11.5342 20.4038 11.5582C20.7804 11.57 21.1341 11.5932 21.4719 11.6156L21.5263 11.6193C21.8195 11.6389 22.1626 11.6618 22.4429 11.6618V7.40825C22.3209 7.40825 22.1219 7.39596 21.7544 7.37149C21.4202 7.34925 20.9976 7.32115 20.5371 7.30672C19.6286 7.27824 18.4672 7.29779 17.3093 7.55377C16.1512 7.8098 14.8404 8.33724 13.8181 9.4452C12.7612 10.5907 12.2266 12.1236 12.2266 13.9619V15.0127H10.6836V19.2662H12.2266V26.6332H16.4801V19.2662H20.3394V15.0127H16.4801V13.9619Z" fill="currentColor"/></svg>from_file</span></h4><!----> <a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.from_file" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.from_file"><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><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/modular_pipelines/pika_h3_flex/references.py#L147" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">media<span class="opacity-60"></span></span></span><!--]--> <span>)</span> <!--[0--><span class="font-bold">→</span> <span class="rounded hover:bg-gray-400 cursor-pointer"><!----><span><code>PikaH3FlexVideoReference</code></span><!----></span><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[0--><p class="flex items-center font-semibold !mt-2 !mb-2 text-gray-800">パラメータ <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700 ml-3"></span></p> <ul class="px-2"><!--[--><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.from_file.media" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.from_file.media"><span><svg class="text-smd" 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><!----><strong>media</strong> (<code>str</code> または <code>os.PathLike</code>) — 動画のパスまたはURL。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[0--><div id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexVideoReference.from_file.returns" class="flex items-center font-semibold space-x-3 text-base !mt-0 !mb-0 text-gray-800 rounded "><p class="text-base"><!---->戻り値<!----></p> <!--[0--><!---->
<p><code>PikaH3FlexVideoReference</code></p>
<!----><!--]--> <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700"></span></div> <div class="text-base"><!---->
<p><code>(num_frames, height, width, 3)</code> の <code>uint8</code> フレームを、コンテナが報告するフレームレートで保持します。音声ストリームがある場合は、そのサウンドトラックおよびそのサウンドトラック自体のサンプルレートも保持します。</p>
<!----></div><!--]--> <!--[-1--><!--]--></div></div><!----> <p><code>PikaH3FlexVideoReference</code> にデコードする際、元の解像度、フレームレート、および音声トラックを保持します。</p> <p>レートは参照に反映されます。これが <a href="/docs/diffusers/main/ja/api/utilities#diffusers.utils.load_video">load_video()</a> を使わずにこの方法でデコードする理由です。MiniMax-H3 は参照を独自の 24 fps にリサンプリングするため、入力時に失われたフレームレートは、誤った速度で条件付けされたリクエストを意味し、それについて警告も出ません。メタデータが誤っているコンテナは、返された参照の <code>fps</code> または <code>sample_rate</code> を上書きすることで修正できます。</p> <p><a href="https://github.com/PyAV-Org/PyAV" rel="nofollow">PyAV</a>が必要です。</p></div></div> <!--[1--><h2 class="relative group"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference" 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference"><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><!---->PikaH3FlexAudioReference<!----></span></h2><!--]--><!----> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference"><!----><h3 class="!m-0"><span class="flex-1 break-all md:text-lg bg-gradient-to-r px-2.5 py-1.5 rounded-xl from-indigo-50/70 to-white dark:from-gray-900 dark:to-gray-950 dark:text-indigo-300 text-indigo-700"><svg class="mr-1.5 text-indigo-500 inline-block -mt-0.5" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" aria-hidden="true" focusable="false" role="img" width=".8em" height=".8em" preserveAspectRatio="xMidYMid meet" viewBox="0 0 24 24"><path class="uim-quaternary" d="M20.23 7.24L12 12L3.77 7.24a1.98 1.98 0 0 1 .7-.71L11 2.76c.62-.35 1.38-.35 2 0l6.53 3.77c.29.173.531.418.7.71z" opacity=".25" fill="currentColor"></path><path class="uim-tertiary" d="M12 12v9.5a2.09 2.09 0 0 1-.91-.21L4.5 17.48a2.003 2.003 0 0 1-1-1.73v-7.5a2.06 2.06 0 0 1 .27-1.01L12 12z" opacity=".5" fill="currentColor"></path><path class="uim-primary" d="M20.5 8.25v7.5a2.003 2.003 0 0 1-1 1.73l-6.62 3.82c-.275.13-.576.198-.88.2V12l8.23-4.76c.175.308.268.656.27 1.01z" fill="currentColor"></path></svg><span class="font-light">class</span> <span class="font-medium">diffusers.modular_pipelines.pika_h3_flex.</span><span class="font-semibold">PikaH3FlexAudioReference</span></span></h3><!----> <a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference"><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><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/modular_pipelines/pika_h3_flex/references.py#L174" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">audio<span class="opacity-60">: Tensor</span></span></span><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">sample_rate<span class="opacity-60">: int | None = None</span></span></span><!--]--> <span>)</span> <!--[-1--><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[0--><p class="flex items-center font-semibold !mt-2 !mb-2 text-gray-800">パラメータ <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700 ml-3"></span></p> <ul class="px-2"><!--[--><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.audio" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.audio"><span><svg class="text-smd" 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><!----><strong>audio</strong> (<code>torch.Tensor</code>、形状 <code>(channels, num_samples)</code>) —
モノラルまたはステレオの参照波形。<!----></span></span></li><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.sample_rate" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.sample_rate"><span><svg class="text-smd" 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><!----><strong>sample_rate</strong> (<code>int</code>, <em>任意</em>) —
<code>audio</code> がサンプルを保持するレートです。省略した場合、オーディオ VAE 自身のレートが使用され、サンプルは変更されません。それ以外のレートが指定された場合は、そのレートにリサンプリングされます。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>音声または音楽のリファレンス: 1リクエストあたり最大3つまで。単独では使用できません。音声リファレンスは、少なくとも1つの画像または動画リファレンスと組み合わせて使用する必要があります。コンディショナーには渡されず、オーディオVAEのみによってエンコードされます。</p> <div class="docstring border-l-2 border-t-2 pl-4 pt-3.5 border-gray-100 rounded-tl-xl mb-6 mt-8"><div><span class="group flex space-x-1.5 items-center text-gray-800 bg-gradient-to-r rounded-tr-lg -mt-4 -ml-4 pt-3 px-2.5" id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.from_file"><!----><h4 class="!m-0"><span class="flex-1 rounded-xl py-0.5 break-all bg-gradient-to-r from-blue-50/60 to-white dark:from-gray-900 dark:to-gray-950 text-blue-700 dark:text-blue-300 font-medium px-2"><svg width="1em" height="1em" viewBox="0 0 32 33" class="mr-1 inline-block -mt-0.5"  xmlns="http://www.w3.org/2000/svg"><path d="M5.80566 18.3545C4.90766 17.4565 4.90766 16.0005 5.80566 15.1025L14.3768 6.53142C15.2748 5.63342 16.7307 5.63342 17.6287 6.53142L26.1999 15.1025C27.0979 16.0005 27.0979 17.4565 26.1999 18.3545L17.6287 26.9256C16.7307 27.8236 15.2748 27.8236 14.3768 26.9256L5.80566 18.3545Z" fill="currentColor" fill-opacity="0.25"/><path fill-rule="evenodd" clip-rule="evenodd" d="M16.4801 13.9619C16.4801 12.9761 16.7467 12.5436 16.9443 12.3296C17.1764 12.078 17.5731 11.8517 18.2275 11.707C18.8821 11.5623 19.638 11.5342 20.4038 11.5582C20.7804 11.57 21.1341 11.5932 21.4719 11.6156L21.5263 11.6193C21.8195 11.6389 22.1626 11.6618 22.4429 11.6618V7.40825C22.3209 7.40825 22.1219 7.39596 21.7544 7.37149C21.4202 7.34925 20.9976 7.32115 20.5371 7.30672C19.6286 7.27824 18.4672 7.29779 17.3093 7.55377C16.1512 7.8098 14.8404 8.33724 13.8181 9.4452C12.7612 10.5907 12.2266 12.1236 12.2266 13.9619V15.0127H10.6836V19.2662H12.2266V26.6332H16.4801V19.2662H20.3394V15.0127H16.4801V13.9619Z" fill="currentColor"/></svg>from_file</span></h4><!----> <a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.from_file" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.from_file"><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><!----></a> <!--[0--><a class="!ml-auto !text-gray-400 !no-underline text-sm flex items-center" href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/modular_pipelines/pika_h3_flex/references.py#L193" target="_blank"><span>&lt;</span> <span class="hidden md:block mx-0.5 hover:!underline">source</span> <span>></span></a><!--]--></span> <!--[0--><p class="font-mono text-xs md:text-sm !leading-relaxed !my-6"><span>(</span> <!--[--><span class="comma cursor-pointer"><span class="rounded hover:bg-black hover:text-white dark:hover:bg-white dark:hover:text-black">media<span class="opacity-60"></span></span></span><!--]--> <span>)</span> <!--[0--><span class="font-bold">→</span> <span class="rounded hover:bg-gray-400 cursor-pointer"><!----><span><code>PikaH3FlexAudioReference</code></span><!----></span><!--]--></p><!--]--> <div class="!mb-10 relative docstring-details "><!--[-1--><!--]--> <!--[0--><p class="flex items-center font-semibold !mt-2 !mb-2 text-gray-800">パラメータ <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700 ml-3"></span></p> <ul class="px-2"><!--[--><li class="text-base !pl-4 my-3 rounded "><span class="group flex space-x-1.5 items-start"><a id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.from_file.media" class="header-link block pr-0.5 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="#diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.from_file.media"><span><svg class="text-smd" 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><!----><strong>media</strong> (<code>str</code> または <code>os.PathLike</code>) — 音声のパスまたは URL、またはサウンドトラックを抽出する動画のパスまたは URL。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[0--><div id="diffusers.modular_pipelines.pika_h3_flex.PikaH3FlexAudioReference.from_file.returns" class="flex items-center font-semibold space-x-3 text-base !mt-0 !mb-0 text-gray-800 rounded "><p class="text-base"><!---->戻り値<!----></p> <!--[0--><!---->
<p><code>PikaH3FlexAudioReference</code></p>
<!----><!--]--> <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700"></span></div> <div class="text-base"><!---->
<p>コンテナが報告するサンプルレートでの、<code>(channels, num_samples)</code> の float32 波形。</p>
<!----></div><!--]--> <!--[-1--><!--]--></div></div><!----> <p>オーディオファイルを、そのファイルが持つサンプルレートで<code>PikaH3FlexAudioReference</code>にデコードします。</p> <p><a href="https://github.com/PyAV-Org/PyAV" rel="nofollow">PyAV</a>が必要です。</p></div></div> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/diffusers/blob/main/docs/source/en/api/pipelines/pika_h3_flex.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, 225],
							data: [null,null],
							form: null,
							error: null
						});
					});
				}
			</script>
		
Xet Storage Details

Size:: 114 kB
Xet hash:: 137e953df394ec8ed5fbbffbd84b37273aa3d07ac587dba7da6f08695f2f3fd8
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.