Buckets:
| <meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"キャッシュ","local":"caching","sections":[{"title":"キャッシュメソッドを選択","local":"choose-a-cache-method","sections":[],"depth":2},{"title":"Pyramid Attention Broadcast","local":"pyramid-attention-broadcast","sections":[],"depth":2},{"title":"FasterCache","local":"fastercache","sections":[],"depth":2},{"title":"SeaCache","local":"seacache","sections":[],"depth":2},{"title":"FirstBlockCache","local":"firstblockcache","sections":[],"depth":2},{"title":"TaylorSeer キャッシュ","local":"taylorseer-cache","sections":[],"depth":2},{"title":"MagCache","local":"magcache","sections":[],"depth":2},{"title":"テキスト KV キャッシュ","local":"text-kv-cache","sections":[],"depth":2},{"title":"新しいモデルにキャッシュを追加する","local":"add-caching-to-a-new-model","sections":[],"depth":2}],"depth":1}"/> | |
| <link href="/docs/diffusers/main/en/_app/immutable/entry/start.B6roO9Zr.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/main/en/_app/immutable/chunks/CjYFLaCl.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/main/en/_app/immutable/chunks/Cc4T7sti.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/main/en/_app/immutable/entry/app.CxxEVG8A.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/main/en/_app/immutable/chunks/C8F28shN.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/main/en/_app/immutable/chunks/BTnLXsn-.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.CmHT0Edy.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/main/en/_app/immutable/chunks/C7Ky3Ped.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/main/en/_app/immutable/nodes/8.DhsC3tNN.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/main/en/_app/immutable/chunks/D2YP2_0L.js" rel="modulepreload"> | |
| <!--fs68f1--><meta name="hf:doc:metadata" content="{"title":"Caching","local":"caching","sections":[{"title":"Choose a cache method","local":"choose-a-cache-method","sections":[],"depth":2},{"title":"Pyramid Attention Broadcast","local":"pyramid-attention-broadcast","sections":[],"depth":2},{"title":"FasterCache","local":"fastercache","sections":[],"depth":2},{"title":"SeaCache","local":"seacache","sections":[],"depth":2},{"title":"FirstBlockCache","local":"firstblockcache","sections":[],"depth":2},{"title":"TaylorSeer Cache","local":"taylorseer-cache","sections":[],"depth":2},{"title":"MagCache","local":"magcache","sections":[],"depth":2},{"title":"Text KV Cache","local":"text-kv-cache","sections":[],"depth":2},{"title":"Add caching to a new model","local":"add-caching-to-a-new-model","sections":[],"depth":2}],"depth":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="caching" 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="#caching"><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><!---->キャッシュ<!----></span></h1><!--]--><!----> <p>Cachingは、デノイジングステップ間で中間層の出力を再利用して推論を高速化します。メモリ使用量が増加しますが、トレーニングは不要です。Cachingをサポートするtransformerで、設定を指定してメソッドを有効にします。</p> <!--[1--><h2 class="relative group"><a id="choose-a-cache-method" 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="#choose-a-cache-method"><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><!---->キャッシュメソッドを選択<!----></span></h2><!--]--><!----> <p>設定する設定項目の量と、必要な精度に応じて、適切な方法を選択してください。各方法は、速度、メモリ使用量、およびキャッシュなしの実行との出力の一致度合いを、それぞれ異なるバランスで調整しています。ある方法に頼る前に、ご自身のプロンプトに対して、キャッシュありとキャッシュなしの出力を比較してください。</p> <table><thead><tr><th>メソッド</th><th>使用場面</th><th>注記</th></tr></thead><tbody><tr><td><a href="#text-kv-cache">テキスト KV キャッシュ</a></td><td>NucleusMoEの画像、ステップ間でテキストのK/V投影を再利用</td><td>現在、NucleusMoEの画像ブロックをフックしています</td></tr><tr><td><a href="#seacache">SeaCache</a></td><td>Cosmos 3 動画生成</td><td>設定はモデル間で引き継がれないことがよくあります</td></tr><tr><td><a href="#firstblockcache">FirstBlockCache</a></td><td>登録済みtransformerに、速度と品質を調整するための1つのメインノブを設けたい</td><td><code>threshold</code> 1つが、ステップをスキップする頻度を制御します</td></tr><tr><td><a href="#magcache">MagCache</a></td><td>checkpointとschedulerのmagnitude ratioが既に存在するか、それとも先にキャリブレーションを行うか</td><td>比率はチェックポイントとスケジューラ固有のものです</td></tr><tr><td><a href="#taylorseer-cache">TaylorSeer</a></td><td>後続のステップの活性化を、先行するステップから予測したい</td><td><code>max_order</code>が大きいほど、メモリ使用量が増加します</td></tr><tr><td><a href="#pyramid-attention-broadcast">PAB</a></td><td>Video、注意の再利用を調整する意図あり(注意の種類ごとにブロックとタイムステップのスキップ範囲)</td><td>スキップ範囲はモデルごとに調整が必要です</td></tr><tr><td><a href="#fastercache">FasterCache</a></td><td>PABと同様に、オプションのCFGブランチのスキッピングを追加</td><td>実験的</td></tr></tbody></table> <!--[1--><h2 class="relative group"><a id="pyramid-attention-broadcast" 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="#pyramid-attention-broadcast"><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><!---->Pyramid Attention Broadcast<!----></span></h2><!--]--><!----> <p><a href="https://huggingface.co/papers/2408.12588" rel="nofollow">Pyramid Attention Broadcast (PAB)</a> は、一部のブロックとタイムステップについて注意の出力を再利用し、すべてのステップで再計算しないことで、デノイジングステップ間の注意を近似します。モデルがそれらを持つ場合、Config は注意の種類(空間、時間、クロス)を分離します。すべての動画モデルが3種類すべてを公開するわけではなく、存在するブロックに一致する範囲のみを設定してください。</p> <p>各タイプは、<code>*_attention_block_skip_range</code>(ウィンドウ内で再計算と再利用を切り替える頻度)と、<code>*_attention_timestep_skip_range</code>(スキップ可能なデノイジングタイムステップ)を使用します。フックがパイプラインの現在のタイムステップを読み取れるように、<code>current_timestep_callback</code> を渡す必要があります。スキップ範囲が広いほど、またはより積極的であるほど、速度は向上しますが、品質リスクも高くなります。</p> <p>有効にするには、<a href="/docs/diffusers/main/ja/api/cache#diffusers.PyramidAttentionBroadcastConfig">PyramidAttentionBroadcastConfig</a> を渡します。</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-python "><!----><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> CogVideoXPipeline, PyramidAttentionBroadcastConfig | |
| pipe = CogVideoXPipeline.from_pretrained(<span class="hljs-string">"THUDM/CogVideoX-5b"</span>, dtype=torch.bfloat16) | |
| pipe.to(<span class="hljs-string">"cuda"</span>) <span class="hljs-comment"># or "mps", "xpu", "cpu"</span> | |
| config = PyramidAttentionBroadcastConfig( | |
| spatial_attention_block_skip_range=<span class="hljs-number">2</span>, | |
| spatial_attention_timestep_skip_range=(<span class="hljs-number">100</span>, <span class="hljs-number">800</span>), | |
| current_timestep_callback=<span class="hljs-keyword">lambda</span>: pipe.current_timestep, | |
| ) | |
| pipe.transformer.enable_cache(config)<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="fastercache" 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="#fastercache"><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><!---->FasterCache<!----></span></h2><!--]--><!----> <p><a href="https://huggingface.co/papers/2410.19355" rel="nofollow">FasterCache</a>は、<a href="#pyramid-attention-broadcast">PAB</a>と同様に、アテンション特徴量をキャッシュして再利用します。また、classifier-free guidanceにおける無条件ブランチをスキップし、連続する潜在変数が十分に冗長である場合、条件付きブランチから推定することもできます。</p> <p>有効にするには、<a href="/docs/diffusers/main/ja/api/cache#diffusers.FasterCacheConfig">FasterCacheConfig</a> を渡します。PABと同様に、使用しているアテンションの種類について<code>*_attention_block_skip_range</code>と<code>*_attention_timestep_skip_range</code>を設定し、必要に応じてCFGブランチのスキップオプションも設定します。</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-python "><!----><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> CogVideoXPipeline, FasterCacheConfig | |
| pipe = CogVideoXPipeline.from_pretrained(<span class="hljs-string">"THUDM/CogVideoX-5b"</span>, dtype=torch.bfloat16) | |
| pipe.to(<span class="hljs-string">"cuda"</span>) <span class="hljs-comment"># or "mps", "xpu", "cpu"</span> | |
| config = FasterCacheConfig( | |
| spatial_attention_block_skip_range=<span class="hljs-number">2</span>, | |
| spatial_attention_timestep_skip_range=(-<span class="hljs-number">1</span>, <span class="hljs-number">681</span>), | |
| current_timestep_callback=<span class="hljs-keyword">lambda</span>: pipe.current_timestep, | |
| attention_weight_callback=<span class="hljs-keyword">lambda</span> _: <span class="hljs-number">0.3</span>, | |
| unconditional_batch_skip_range=<span class="hljs-number">5</span>, | |
| unconditional_batch_timestep_skip_range=(-<span class="hljs-number">1</span>, <span class="hljs-number">781</span>), | |
| tensor_format=<span class="hljs-string">"BFCHW"</span>, | |
| ) | |
| pipe.transformer.enable_cache(config)<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="seacache" 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="#seacache"><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><!---->SeaCache<!----></span></h2><!--]--><!----> <p><a href="https://huggingface.co/papers/2602.18993" rel="nofollow">SeaCache</a>は、連続するノイズ除去ステップ間でSpectral Evolution Aware (SEA)指標を比較します。累積変化が閾値を下回る場合、transformerブロックスタックをスキップし、キャッシュされた残差から出力を予測します。この手法は近似であり、動画生成向けに設計されています。</p> <p>SeaCacheの組み込みアダプターには、以下が含まれます:</p> <ul><li>Cosmos 3は、主に最適化およびベンチマークされた統合です。</li> <li>Wan T2Vは、SeaCacheに生ビジョン潜在変数をどのように提供するかを示すデモとして、汎用的な繰り返しブロックパスを使用しています。同じキャッシュパラメータがWanやその他のWanバリアントにそのまま適用できるわけではありません。</li></ul> <p>transformer で SeaCache を有効にします。Cosmos 3 のデノイジングループは、各 <code>cache_context</code> にスケジューラーのステップ、シグマ、ステップ数を付与するため、追加のパラメータは不要です。</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-python "><!----><span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> Cosmos3OmniPipeline, SeaCacheConfig | |
| pipe = Cosmos3OmniPipeline.from_pretrained(<span class="hljs-string">"nvidia/Cosmos3-Nano"</span>) | |
| pipe.transformer.enable_cache(SeaCacheConfig(threshold=<span class="hljs-number">0.2</span>, max_consecutive_cached=<span class="hljs-number">2</span>))<!----></pre></div><!----> <p>SeaCache は出力を変更する可能性があります。すべてのステップで完全な transformer を実行する必要がある場合は、<code>pipe.transformer.disable_cache()</code> を呼び出してください。同じ有効化の呼び出しは、<a href="/docs/diffusers/main/ja/api/pipelines/cosmos3#diffusers.Cosmos3OmniPipeline">Cosmos3OmniPipeline</a>、<a href="/docs/diffusers/main/ja/api/pipelines/cosmos3#diffusers.Cosmos3OmniModularPipeline">Cosmos3OmniModularPipeline</a>、および <a href="/docs/diffusers/main/ja/api/pipelines/cosmos3#diffusers.Cosmos3DistilledModularPipeline">Cosmos3DistilledModularPipeline</a> でも機能します。</p> <p>別の動画 Transformer で SeaCache を使用するには、<a href="#add-caching-to-a-new-model">新しいモデルにキャッシュを追加する</a>の手順に従ってください。SeaCache には、すべての <code>cache_context</code> で <code>step_index</code>、<code>sigma</code>、<code>num_inference_steps</code> が必要であり、組み込みのアダプターが存在しない場合は <code>raw_vision_callback</code> も必要です。モデルとスケジューラーごとにパラメータを調整してください。</p> <!--[1--><h2 class="relative group"><a id="firstblockcache" 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="#firstblockcache"><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><!---->FirstBlockCache<!----></span></h2><!--]--><!----> <p><a href="/docs/diffusers/main/ja/api/cache#diffusers.FirstBlockCacheConfig">FirstBlockCacheConfig</a> は、denoiser の初期レイヤーがタイムステップごとにどの程度変化するかを確認します。変化が小さい場合、モデルはコストの高い後続のレイヤーをスキップし、前回の出力を再利用します。</p> <p><code>enable_cache</code> を使用して有効化し、<code>disable_cache</code> と <code>is_cache_enabled</code> が同期した状態を維持してください。デフォルトの <code>threshold</code> は <code>0.05</code> です。<code>0.2</code> のような高い値を設定すると、より頻繁にスキップされるため速度が向上しますが、生成品質が低下する可能性があります。</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-python "><!----><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> DiffusionPipeline, FirstBlockCacheConfig | |
| pipe = DiffusionPipeline.from_pretrained( | |
| <span class="hljs-string">"Qwen/Qwen-Image"</span>, dtype=torch.bfloat16 | |
| ) | |
| pipe.transformer.enable_cache(FirstBlockCacheConfig(threshold=<span class="hljs-number">0.2</span>))<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="taylorseer-cache" 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="#taylorseer-cache"><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><!---->TaylorSeer キャッシュ<!----></span></h2><!--]--><!----> <p><a href="https://huggingface.co/papers/2503.06923" rel="nofollow">TaylorSeer Cache</a>は、デノイジングステップにわたるテイラー級数展開を用いて拡散モデルの推論を高速化します。初期ステップの活性化から後続ステップの活性化を予測し、その予測を複数ステップにわたって再利用することで、Transformerの完全な計算量を削減します。</p> <ul><li><code>cache_interval</code>: 完全なフォワードパスを実行する前に、キャッシュされた出力を再利用するステップ数</li> <li><code>disable_cache_before_step</code>: 近似のためのデータを収集するために、完全な計算を行う初期ステップ</li> <li><code>max_order</code>: Taylor の次数を高くすると精度が向上する場合がありますが、メモリ使用量が増えます。変更する理由がない限り、<code>1</code> のままにしてください。</li></ul> <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-python "><!----><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> FluxPipeline, TaylorSeerCacheConfig | |
| pipe = FluxPipeline.from_pretrained( | |
| <span class="hljs-string">"black-forest-labs/FLUX.1-dev"</span>, | |
| dtype=torch.bfloat16, | |
| ).to(<span class="hljs-string">"cuda"</span>) <span class="hljs-comment"># or "mps", "xpu", "cpu"</span> | |
| config = TaylorSeerCacheConfig( | |
| cache_interval=<span class="hljs-number">5</span>, | |
| max_order=<span class="hljs-number">1</span>, | |
| disable_cache_before_step=<span class="hljs-number">10</span>, | |
| taylor_factors_dtype=torch.bfloat16, | |
| ) | |
| pipe.transformer.enable_cache(config)<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="magcache" 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="#magcache"><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><!---->MagCache<!----></span></h2><!--]--><!----> <p><a href="https://github.com/Zehong-Ma/MagCache" rel="nofollow">MagCache</a>は、残差更新の大きさに基づいてtransformerブロックをスキップします。更新の大きさは、デノイジングの過程で予測可能な形で減衰し、MagCacheは事前計算された大きさの比率(<code>mag_ratios</code>)から誤差予算を追跡することで、再利用が安全なタイミングを判断します。これらの比率はチェックポイントとスケジューラに依存します。ステップ数が大きい場合の比率は、ステップ数が少ない場合に補間して適用できます。</p> <blockquote class="tip"><p>バンドルされている<code>FLUX_MAG_RATIOS</code>はFLUX.1で測定されたものです。Flux2やFlux2 Kleinを含む他のモデルには、それぞれ独自のキャリブレーション実行が必要です。</p></blockquote> <p>MagCacheは2つのステップに従います:</p> <ol><li>キャリブレーション: <code>calibrate=True</code>を指定して推論を1回実行します。フックが残差の大きさを測定し、計算された比率を出力します。</li> <li>推論: キャリブレーションキャッシュを無効化し、その後、これらの比率を<code>MagCacheConfig</code>に渡して高速化します。</li></ol> <p>Classifier-free guidance はキャリブレーションに影響を与える可能性があります。Flux のように <code>true_cfg_scale > 1</code> の場合に true CFG とシーケンシャルコンテキストを使用するパイプラインでは、<code>cache_context("cond")</code> と <code>cache_context("uncond")</code> がそれぞれ別々に呼び出されます。キャリブレーションではコンテキストごとに1つの配列が出力される場合がありますが、ほとんどの場合は条件付きの配列を使用してください。条件付き入力と非条件付き入力を連結して CFG をバッチ処理するパイプライン(CogVideoX など)では、そのまま直接使用できる単一の結合配列が生成されます。</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-python "><!----><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> FluxPipeline, MagCacheConfig | |
| <span class="hljs-keyword">from</span> diffusers.hooks.mag_cache <span class="hljs-keyword">import</span> FLUX_MAG_RATIOS | |
| pipe = FluxPipeline.from_pretrained( | |
| <span class="hljs-string">"black-forest-labs/FLUX.1-schnell"</span>, | |
| dtype=torch.bfloat16 | |
| ).to(<span class="hljs-string">"cuda"</span>) <span class="hljs-comment"># or "mps", "xpu", "cpu"</span> | |
| <span class="hljs-comment"># 1. Calibration Step</span> | |
| <span class="hljs-comment"># Run full inference to measure model behavior.</span> | |
| calib_config = MagCacheConfig(calibrate=<span class="hljs-literal">True</span>, num_inference_steps=<span class="hljs-number">4</span>) | |
| pipe.transformer.enable_cache(calib_config) | |
| <span class="hljs-comment"># Run a prompt to trigger calibration</span> | |
| pipe(<span class="hljs-string">"A cat playing chess"</span>, num_inference_steps=<span class="hljs-number">4</span>) | |
| <span class="hljs-comment"># Prints: [MagCache] Calibration Complete. Copy these values to MagCacheConfig(mag_ratios=...):</span> | |
| <span class="hljs-comment"># 2. Inference Step</span> | |
| <span class="hljs-comment"># Disable calibration hooks before enabling MagCache for inference.</span> | |
| pipe.transformer.disable_cache() | |
| <span class="hljs-comment"># Apply ratios from calibration, or use the Flux defaults:</span> | |
| <span class="hljs-comment"># mag_ratios=FLUX_MAG_RATIOS</span> | |
| mag_config = MagCacheConfig( | |
| mag_ratios=[<span class="hljs-number">1.0</span>, <span class="hljs-number">1.37</span>, <span class="hljs-number">0.97</span>, <span class="hljs-number">0.87</span>], | |
| num_inference_steps=<span class="hljs-number">4</span> | |
| ) | |
| pipe.transformer.enable_cache(mag_config) | |
| image = pipe(<span class="hljs-string">"A cat playing chess"</span>, num_inference_steps=<span class="hljs-number">4</span>).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="text-kv-cache" class="header-link block pr-1.5 text-lg no-hover:hidden with-hover:absolute with-hover:p-1.5 with-hover:opacity-0 with-hover:group-hover:opacity-100 with-hover:right-full" href="#text-kv-cache"><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><!---->テキスト KV キャッシュ<!----></span></h2><!--]--><!----> <p><a href="/docs/diffusers/main/ja/api/cache#diffusers.TextKVCacheConfig">TextKVCacheConfig</a> は、テキストのキーとバリューの射影を一度だけ計算し、再計算せずにすべてのデノイジングステップで再利用します。現在、<a href="/docs/diffusers/main/ja/api/cache#diffusers.apply_text_kv_cache">apply_text_kv_cache()</a> は <code>NucleusMoEImageTransformerBlock</code> に対してのみフックされています。<code>enable_cache</code> を有効にして有効化してください。</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-python "><!----><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> NucleusMoEImagePipeline, TextKVCacheConfig | |
| pipe = NucleusMoEImagePipeline.from_pretrained( | |
| <span class="hljs-string">"NucleusAI/NucleusMoE-Image"</span>, dtype=torch.bfloat16 | |
| ) | |
| pipe.to(<span class="hljs-string">"cuda"</span>) <span class="hljs-comment"># or "mps", "xpu", "cpu"</span> | |
| pipe.transformer.enable_cache(TextKVCacheConfig()) | |
| image = pipe(<span class="hljs-string">"A cat holding a sign that says hello world"</span>, num_inference_steps=<span class="hljs-number">50</span>).images[<span class="hljs-number">0</span>]<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="add-caching-to-a-new-model" 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="#add-caching-to-a-new-model"><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><!---->新しいモデルにキャッシュを追加する<!----></span></h2><!--]--><!----> <p>Cacheメソッドはフックを介してモデルに接続され、これらのフックはモデルクラス、そのtransformerブロック、およびパイプラインのdenoisingループに依存します。<code>enable_cache</code>を新しいモデルに呼び出す前に、これら3つすべてを設定してください。</p> <ol><li><p>transformer クラスで <a href="/docs/diffusers/main/ja/api/cache#diffusers.CacheMixin">CacheMixin</a> を継承します。これにより、<code>enable_cache</code>、<code>disable_cache</code>、<code>cache_context</code> が追加されます。</p></li> <li><p><code>hooks/_helpers.py</code> の <code>_register_transformer_blocks_metadata</code> に、transformer ブロックの出力レイアウトを登録します。FirstBlockCache、MagCache、SeaCache など、ブロック全体をスキップまたは再利用するメソッドは、ブロックの出力から隠れ状態を見つけるためにこれを使用します。</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 "><!---->TransformerBlockRegistry.register( | |
| model_class=MyTransformerBlock, | |
| metadata=TransformerBlockMetadata( | |
| return_hidden_states_index=<span class="hljs-number">0</span>, | |
| return_encoder_hidden_states_index=<span class="hljs-literal">None</span>, | |
| ), | |
| )<!----></pre></div><!----></li> <li><p>パイプラインのデノイジングループ内の各transformer呼び出しを<code>cache_context</code>でラップします。classifier-free guidanceが条件付きと条件なしのブランチを個別の呼び出しとして実行する場合、<code>"cond"</code>や<code>"uncond"</code>のような別々の名前を使用し、各ブランチが独自のキャッシュ状態を保持できるようにします。</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">with</span> <span class="hljs-variable language_">self</span>.transformer.cache_context(<span class="hljs-string">"cond"</span>): | |
| noise_pred = <span class="hljs-variable language_">self</span>.transformer(hidden_states=latents, timestep=timestep, ...)[<span class="hljs-number">0</span>]<!----></pre></div><!----></li></ol> <p>PABおよびFasterCacheも、フックが現在のタイムステップを読み取れるように<code>current_timestep_callback</code>が必要です。新しいモデルでの出力品質を確認してください。1つのモデル用に調整されたキャッシュ設定は、別のモデルには適用されないことが多いためです。</p> <a class="!text-gray-400 !no-underline text-sm flex items-center not-prose mt-4" href="https://github.com/huggingface/diffusers/blob/main/docs/source/en/optimization/cache.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_kft2r9 = { | |
| 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.B6roO9Zr.js"), | |
| import("/docs/diffusers/main/en/_app/immutable/entry/app.CxxEVG8A.js") | |
| ]).then(([kit, app]) => { | |
| kit.start(app, element, { | |
| node_ids: [0, 8], | |
| data: [null,null], | |
| form: null, | |
| error: null | |
| }); | |
| }); | |
| } | |
| </script> | |
Xet Storage Details
- Size:
- 50.3 kB
- Xet hash:
- b8f91d83b322dd418cecf25d68647e8deb2a61084ba57a3be6cca69f406ffcc5
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.