Buckets:
| <meta charset="utf-8" /><meta name="hf:doc:metadata" content="{"title":"Framepack","local":"framepack","sections":[{"title":"可用模型","local":"available-models","sections":[],"depth":2},{"title":"用法","local":"usage","sections":[{"title":"首帧和尾帧转视频","local":"first-and-last-frame-to-video","sections":[],"depth":3},{"title":"基础采样","local":"vanilla-sampling","sections":[],"depth":3},{"title":"组卸载","local":"group-offloading","sections":[],"depth":3}],"depth":2},{"title":"HunyuanVideoFramepackPipeline","local":"diffusers.HunyuanVideoFramepackPipeline","sections":[{"title":"图像到视频","local":"image-to-video","sections":[],"depth":5},{"title":"首帧和尾帧图像转视频","local":"first-and-last-image-to-video","sections":[],"depth":5}],"depth":2},{"title":"HunyuanVideoPipelineOutput","local":"diffusers.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput","sections":[],"depth":2}],"depth":1}"/> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/entry/start.1gaXe03k.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/BObs6oQO.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/DPnyVmXs.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/entry/app.BwdYRbCn.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/DSGUT_OW.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/C045MWWx.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/6tuJ3nyi.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/DsnmJJEf.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/nodes/0.CornNeQR.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/IaDvMxLQ.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/nodes/181.z7c3tQ9A.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/bKdwVYpU.js" rel="modulepreload"> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/chunks/D0iXbNom.js" rel="modulepreload"> | |
| <!--1h3mats--><meta name="hf:doc:metadata" content="{"title":"Framepack","local":"framepack","sections":[{"title":"Available models","local":"available-models","sections":[],"depth":2},{"title":"Usage","local":"usage","sections":[{"title":"First and last frame to video","local":"first-and-last-frame-to-video","sections":[],"depth":3},{"title":"Vanilla sampling","local":"vanilla-sampling","sections":[],"depth":3},{"title":"Group offloading","local":"group-offloading","sections":[],"depth":3}],"depth":2},{"title":"HunyuanVideoFramepackPipeline","local":"diffusers.HunyuanVideoFramepackPipeline","sections":[{"title":"Image-to-Video","local":"image-to-video","sections":[],"depth":5},{"title":"First and Last Image-to-Video","local":"first-and-last-image-to-video","sections":[],"depth":5}],"depth":2},{"title":"HunyuanVideoPipelineOutput","local":"diffusers.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput","sections":[],"depth":2}],"depth":1}"/><!----> | |
| <link href="/docs/diffusers/v0.43.0/en/_app/immutable/assets/0.tn0RQdqM.css" rel="modulepreload"> <!--[--><!--[0--><!--[--><!--[0--><!--[--><p></p> <!--[0--><h1 class="relative group"><a id="framepack" 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="#framepack"><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><!---->Framepack<!----></span></h1><!--]--><!----> <div class="flex flex-wrap space-x-1"><img alt="LoRA" src="https://img.shields.io/badge/LoRA-d8b4fe?style=flat"/></div> <p>Lvmin Zhang 和 Maneesh Agrawala 的《<a href="https://huggingface.co/papers/2504.12626" rel="nofollow">Packing Input Frame Context in Next-Frame Prediction Models for Video Generation</a>》</p> <p><em>我们提出了一种名为 FramePack 的神经网络结构，用于训练视频生成的下一帧（或下一帧片段）预测模型。FramePack 对输入帧进行压缩，使得无论视频长度如何，Transformer 的上下文长度都保持为一个固定值。因此，我们能够使用视频扩散模型处理大量帧，其计算瓶颈与图像扩散模型相当。这也使得训练时的视频批次大小显著提高（批次大小可与图像扩散训练相媲美）。我们还提出了一种抗漂移采样方法，该方法以倒序时间顺序生成帧，并预先建立端点，以避免暴露偏差（即迭代过程中的误差累积）。最后，我们展示了现有的视频扩散模型可以通过 FramePack 进行微调，并且由于下一帧预测支持更均衡的扩散调度器以及更极端的流形时间步，其视觉质量可能会得到提升。</em></p> <blockquote class="tip"><p>请务必参阅 <a href="../../using-diffusers/schedulers">调度器指南</a>，了解如何探索调度器速度与质量之间的权衡，并查看 <a href="../../using-diffusers/loading#reusing-models-in-multiple-pipelines">跨流水线复用组件</a> 部分，了解如何高效地将相同组件加载到多个流水线中。</p></blockquote> <!--[1--><h2 class="relative group"><a id="available-models" 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="#available-models"><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><!---->可用模型<!----></span></h2><!--]--><!----> <table><thead><tr><th align="left">模型名称</th><th align="left">描述</th></tr></thead><tbody><tr><td align="left">- <a href="https://huggingface.co/lllyasviel/FramePackI2V_HY" rel="nofollow"><code>lllyasviel/FramePackI2V_HY</code></a></td><td align="left">采用论文中描述的“反转抗漂移”策略进行训练。在运行流水线时，推理需要设置 <code>sampling_type="inverted_anti_drifting"</code>。</td></tr><tr><td align="left">- <a href="https://huggingface.co/lllyasviel/FramePack_F1_I2V_HY_20250503" rel="nofollow"><code>lllyasviel/FramePack_F1_I2V_HY_20250503</code></a></td><td align="left">该模型采用一种新颖的抗漂移策略进行训练，但推理过程则按照论文中所述采用“vanilla”策略执行。在运行 pipeline 时，需要将 <code>sampling_type="vanilla"</code> 设置为该值以进行推理。</td></tr></tbody></table> <!--[1--><h2 class="relative group"><a id="usage" 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="#usage"><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><!---->用法<!----></span></h2><!--]--><!----> <p>请参阅 pipeline 文档以了解基本用法示例。以下部分包含有关 offloading、不同采样方法、量化等的示例。</p> <!--[2--><h3 class="relative group"><a id="first-and-last-frame-to-video" 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="#first-and-last-frame-to-video"><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><!---->首帧和尾帧转视频<!----></span></h3><!--]--><!----> <p>以下示例展示了如何使用 Framepack 配合起始和结束图像控制，并采用反向抗漂移采样模型。</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> 已复制</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> HunyuanVideoFramepackPipeline, HunyuanVideoFramepackTransformer3DModel | |
| <span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> export_to_video, load_image | |
| <span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> SiglipImageProcessor, SiglipVisionModel | |
| transformer = HunyuanVideoFramepackTransformer3DModel.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/FramePackI2V_HY"</span>, dtype=torch.bfloat16 | |
| ) | |
| feature_extractor = SiglipImageProcessor.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"feature_extractor"</span> | |
| ) | |
| image_encoder = SiglipVisionModel.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"image_encoder"</span>, dtype=torch.float16 | |
| ) | |
| pipe = HunyuanVideoFramepackPipeline.from_pretrained( | |
| <span class="hljs-string">"hunyuanvideo-community/HunyuanVideo"</span>, | |
| transformer=transformer, | |
| feature_extractor=feature_extractor, | |
| image_encoder=image_encoder, | |
| dtype=torch.float16, | |
| ) | |
| <span class="hljs-comment"># Enable memory optimizations</span> | |
| pipe.enable_model_cpu_offload() | |
| pipe.vae.enable_tiling() | |
| prompt = <span class="hljs-string">"CG animation style, a small blue bird takes off from the ground, flapping its wings. The bird's feathers are delicate, with a unique pattern on its chest. The background shows a blue sky with white clouds under bright sunshine. The camera follows the bird upward, capturing its flight and the vastness of the sky from a close-up, low-angle perspective."</span> | |
| first_image = load_image( | |
| <span class="hljs-string">"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/flf2v_input_first_frame.png"</span> | |
| ) | |
| last_image = load_image( | |
| <span class="hljs-string">"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/flf2v_input_last_frame.png"</span> | |
| ) | |
| output = pipe( | |
| image=first_image, | |
| last_image=last_image, | |
| prompt=prompt, | |
| height=<span class="hljs-number">512</span>, | |
| width=<span class="hljs-number">512</span>, | |
| num_frames=<span class="hljs-number">91</span>, | |
| num_inference_steps=<span class="hljs-number">30</span>, | |
| guidance_scale=<span class="hljs-number">9.0</span>, | |
| generator=torch.Generator().manual_seed(<span class="hljs-number">0</span>), | |
| sampling_type=<span class="hljs-string">"inverted_anti_drifting"</span>, | |
| ).frames[<span class="hljs-number">0</span>] | |
| export_to_video(output, <span class="hljs-string">"output.mp4"</span>, fps=<span class="hljs-number">30</span>)<!----></pre></div><!----> <!--[2--><h3 class="relative group"><a id="vanilla-sampling" 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="#vanilla-sampling"><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><!---->基础采样<!----></span></h3><!--]--><!----> <p>以下示例展示了如何使用 Framepack 与使用常规采样训练、但采用新的抗漂移调节方法的 F1 模型。</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> 已复制</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> HunyuanVideoFramepackPipeline, HunyuanVideoFramepackTransformer3DModel | |
| <span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> export_to_video, load_image | |
| <span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> SiglipImageProcessor, SiglipVisionModel | |
| transformer = HunyuanVideoFramepackTransformer3DModel.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/FramePack_F1_I2V_HY_20250503"</span>, dtype=torch.bfloat16 | |
| ) | |
| feature_extractor = SiglipImageProcessor.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"feature_extractor"</span> | |
| ) | |
| image_encoder = SiglipVisionModel.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"image_encoder"</span>, dtype=torch.float16 | |
| ) | |
| pipe = HunyuanVideoFramepackPipeline.from_pretrained( | |
| <span class="hljs-string">"hunyuanvideo-community/HunyuanVideo"</span>, | |
| transformer=transformer, | |
| feature_extractor=feature_extractor, | |
| image_encoder=image_encoder, | |
| dtype=torch.float16, | |
| ) | |
| <span class="hljs-comment"># Enable memory optimizations</span> | |
| pipe.enable_model_cpu_offload() | |
| pipe.vae.enable_tiling() | |
| image = load_image( | |
| <span class="hljs-string">"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/penguin.png"</span> | |
| ) | |
| output = pipe( | |
| image=image, | |
| prompt=<span class="hljs-string">"A penguin dancing in the snow"</span>, | |
| height=<span class="hljs-number">832</span>, | |
| width=<span class="hljs-number">480</span>, | |
| num_frames=<span class="hljs-number">91</span>, | |
| num_inference_steps=<span class="hljs-number">30</span>, | |
| guidance_scale=<span class="hljs-number">9.0</span>, | |
| generator=torch.Generator().manual_seed(<span class="hljs-number">0</span>), | |
| sampling_type=<span class="hljs-string">"vanilla"</span>, | |
| ).frames[<span class="hljs-number">0</span>] | |
| export_to_video(output, <span class="hljs-string">"output.mp4"</span>, fps=<span class="hljs-number">30</span>)<!----></pre></div><!----> <!--[2--><h3 class="relative group"><a id="group-offloading" 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="#group-offloading"><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><!---->组卸载<!----></span></h3><!--]--><!----> <p>组卸载（<a href="/docs/diffusers/v0.43.0/zh/api/utilities#diffusers.hooks.apply_group_offloading">apply_group_offloading()</a>）为将任意模型的内部组件卸载到 CPU 提供了激进的内存优化，且可能不会给生成时间带来额外开销。如果您的可用 VRAM 非常有限，那么根据可用的 CPU RAM 数量，这种方法可能适合您。</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> 已复制</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> HunyuanVideoFramepackPipeline, HunyuanVideoFramepackTransformer3DModel | |
| <span class="hljs-keyword">from</span> diffusers.hooks <span class="hljs-keyword">import</span> apply_group_offloading | |
| <span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> export_to_video, load_image | |
| <span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> SiglipImageProcessor, SiglipVisionModel | |
| transformer = HunyuanVideoFramepackTransformer3DModel.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/FramePack_F1_I2V_HY_20250503"</span>, dtype=torch.bfloat16 | |
| ) | |
| feature_extractor = SiglipImageProcessor.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"feature_extractor"</span> | |
| ) | |
| image_encoder = SiglipVisionModel.from_pretrained( | |
| <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"image_encoder"</span>, dtype=torch.float16 | |
| ) | |
| pipe = HunyuanVideoFramepackPipeline.from_pretrained( | |
| <span class="hljs-string">"hunyuanvideo-community/HunyuanVideo"</span>, | |
| transformer=transformer, | |
| feature_extractor=feature_extractor, | |
| image_encoder=image_encoder, | |
| dtype=torch.float16, | |
| ) | |
| <span class="hljs-comment"># Enable group offloading</span> | |
| onload_device = torch.device(<span class="hljs-string">"cuda"</span>) | |
| offload_device = torch.device(<span class="hljs-string">"cpu"</span>) | |
| <span class="hljs-built_in">list</span>(<span class="hljs-built_in">map</span>( | |
| <span class="hljs-keyword">lambda</span> x: apply_group_offloading(x, onload_device, offload_device, offload_type=<span class="hljs-string">"leaf_level"</span>, use_stream=<span class="hljs-literal">True</span>, low_cpu_mem_usage=<span class="hljs-literal">True</span>), | |
| [pipe.text_encoder, pipe.text_encoder_2, pipe.transformer] | |
| )) | |
| pipe.image_encoder.to(onload_device) | |
| pipe.vae.to(onload_device) | |
| pipe.vae.enable_tiling() | |
| image = load_image( | |
| <span class="hljs-string">"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/penguin.png"</span> | |
| ) | |
| output = pipe( | |
| image=image, | |
| prompt=<span class="hljs-string">"A penguin dancing in the snow"</span>, | |
| height=<span class="hljs-number">832</span>, | |
| width=<span class="hljs-number">480</span>, | |
| num_frames=<span class="hljs-number">91</span>, | |
| num_inference_steps=<span class="hljs-number">30</span>, | |
| guidance_scale=<span class="hljs-number">9.0</span>, | |
| generator=torch.Generator().manual_seed(<span class="hljs-number">0</span>), | |
| sampling_type=<span class="hljs-string">"vanilla"</span>, | |
| ).frames[<span class="hljs-number">0</span>] | |
| <span class="hljs-built_in">print</span>(<span class="hljs-string">f"Max memory: <span class="hljs-subst">{torch.cuda.max_memory_allocated() / <span class="hljs-number">1024</span>**<span class="hljs-number">3</span>:<span class="hljs-number">.3</span>f}</span> GB"</span>) | |
| export_to_video(output, <span class="hljs-string">"output.mp4"</span>, fps=<span class="hljs-number">30</span>)<!----></pre></div><!----> <!--[1--><h2 class="relative group"><a id="diffusers.HunyuanVideoFramepackPipeline" 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.HunyuanVideoFramepackPipeline"><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><!---->HunyuanVideoFramepackPipeline<!----></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.HunyuanVideoFramepackPipeline"><!----><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">类</span> <span class="font-medium">diffusers.</span><span class="font-semibold">HunyuanVideoFramepackPipeline</span></span></h3><!----> <a id="diffusers.HunyuanVideoFramepackPipeline" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.HunyuanVideoFramepackPipeline"><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/v0.43.0/src/diffusers/pipelines/hunyuan_video/pipeline_hunyuan_video_framepack.py#L243" target="_blank"><span><</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">text_encoder<span class="opacity-60">: LlamaModel</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">tokenizer<span class="opacity-60">: LlamaTokenizer</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">transformer<span class="opacity-60">: HunyuanVideoFramepackTransformer3DModel</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">vae<span class="opacity-60">: AutoencoderKLHunyuanVideo</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">scheduler<span class="opacity-60">: FlowMatchEulerDiscreteScheduler</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">text_encoder_2<span class="opacity-60">: CLIPTextModel</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">tokenizer_2<span class="opacity-60">: CLIPTokenizer</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">image_encoder<span class="opacity-60">: SiglipVisionModel</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">feature_extractor<span class="opacity-60">: SiglipImageProcessorPil</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">参数 <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.HunyuanVideoFramepackPipeline.text_encoder" 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.HunyuanVideoFramepackPipeline.text_encoder"><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>text_encoder</strong> (<code>LlamaModel</code>) — | |
| <a href="https://huggingface.co/xtuner/llava-llama-3-8b-v1_1-transformers" rel="nofollow">Llava Llama3-8B</a>。<!----></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.HunyuanVideoFramepackPipeline.tokenizer" 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.HunyuanVideoFramepackPipeline.tokenizer"><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>tokenizer</strong> (<code>LlamaTokenizer</code>) — | |
| 来自 <a href="https://huggingface.co/xtuner/llava-llama-3-8b-v1_1-transformers" rel="nofollow">Llava Llama3-8B</a> 的分词器。<!----></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.HunyuanVideoFramepackPipeline.transformer" 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.HunyuanVideoFramepackPipeline.transformer"><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>transformer</strong> (<a href="/docs/diffusers/v0.43.0/zh/api/models/hunyuan_video_transformer_3d#diffusers.HunyuanVideoTransformer3DModel">HunyuanVideoTransformer3DModel</a>) — | |
| 用于对编码后的图像潜在变量进行去噪的条件 Transformer。<!----></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.HunyuanVideoFramepackPipeline.scheduler" 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.HunyuanVideoFramepackPipeline.scheduler"><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>scheduler</strong> (<a href="/docs/diffusers/v0.43.0/zh/api/schedulers/flow_match_euler_discrete#diffusers.FlowMatchEulerDiscreteScheduler">FlowMatchEulerDiscreteScheduler</a>) — | |
| 用于与 <code>transformer</code> 配合，对编码后的图像潜变量进行去噪的调度器。<!----></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.HunyuanVideoFramepackPipeline.vae" 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.HunyuanVideoFramepackPipeline.vae"><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>vae</strong> (<a href="/docs/diffusers/v0.43.0/zh/api/models/autoencoder_kl_hunyuan_video#diffusers.AutoencoderKLHunyuanVideo">AutoencoderKLHunyuanVideo</a>) — | |
| 变分自编码器（VAE）模型，用于将视频编码为潜在表示，以及从潜在表示解码为视频。<!----></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.HunyuanVideoFramepackPipeline.text_encoder_2" 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.HunyuanVideoFramepackPipeline.text_encoder_2"><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>text_encoder_2</strong> (<code>CLIPTextModel</code>) — | |
| <a href="https://huggingface.co/docs/transformers/model_doc/clip#transformers.CLIPTextModel" rel="nofollow">CLIP</a>，具体为 | |
| <a href="https://huggingface.co/openai/clip-vit-large-patch14" rel="nofollow">clip-vit-large-patch14</a> 变体。<!----></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.HunyuanVideoFramepackPipeline.tokenizer_2" 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.HunyuanVideoFramepackPipeline.tokenizer_2"><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>tokenizer_2</strong> (<code>CLIPTokenizer</code>) — | |
| <a href="https://huggingface.co/docs/transformers/en/model_doc/clip#transformers.CLIPTokenizer" rel="nofollow">CLIPTokenizer</a> 类的分词器。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>用于使用 HunyuanVideo 进行文本到视频生成的 Pipeline。</p> <p>该模型继承自 <a href="/docs/diffusers/v0.43.0/zh/api/pipelines/overview#diffusers.DiffusionPipeline">DiffusionPipeline</a>。有关为所有 pipeline 实现的通用方法（下载、保存、在特定设备上运行等），请参阅父类文档。</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.HunyuanVideoFramepackPipeline.__call__"><!----><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>__call__</span></h4><!----> <a id="diffusers.HunyuanVideoFramepackPipeline.__call__" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.HunyuanVideoFramepackPipeline.__call__"><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/v0.43.0/src/diffusers/pipelines/hunyuan_video/pipeline_hunyuan_video_framepack.py#L586" target="_blank"><span><</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, list[PIL.Image.Image], list[numpy.ndarray], list[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">last_image<span class="opacity-60">: typing.Union[PIL.Image.Image, numpy.ndarray, torch.Tensor, list[PIL.Image.Image], list[numpy.ndarray], list[torch.Tensor], NoneType] = 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">prompt<span class="opacity-60">: str | list[str] = 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">prompt_2<span class="opacity-60">: str | list[str] = 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">negative_prompt<span class="opacity-60">: str | list[str] = 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">negative_prompt_2<span class="opacity-60">: str | list[str] = 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">height<span class="opacity-60">: int = 720</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">width<span class="opacity-60">: int = 1280</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">num_frames<span class="opacity-60">: int = 129</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">latent_window_size<span class="opacity-60">: int = 9</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">num_inference_steps<span class="opacity-60">: int = 50</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">sigmas<span class="opacity-60">: list[float] = 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">true_cfg_scale<span class="opacity-60">: float = 1.0</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">guidance_scale<span class="opacity-60">: float = 6.0</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">num_videos_per_prompt<span class="opacity-60">: int | None = 1</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">generator<span class="opacity-60">: typing.Union[torch.Generator, list[torch.Generator], NoneType] = 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">image_latents<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">last_image_latents<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">prompt_embeds<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">pooled_prompt_embeds<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">prompt_attention_mask<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">negative_prompt_embeds<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">negative_pooled_prompt_embeds<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">negative_prompt_attention_mask<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">output_type<span class="opacity-60">: str | None = 'pil'</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">return_dict<span class="opacity-60">: bool = True</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">attention_kwargs<span class="opacity-60">: dict[str, typing.Any] | 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">callback_on_step_end<span class="opacity-60">: typing.Union[typing.Callable[[int, int], NoneType], diffusers.callbacks.PipelineCallback, diffusers.callbacks.MultiPipelineCallbacks, NoneType] = 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">callback_on_step_end_tensor_inputs<span class="opacity-60">: list[str] = ['latents']</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">prompt_template<span class="opacity-60">: dict[str, typing.Any] = {'template': '<|start_header_id|>system<|end_header_id|>\n\nDescribe the video by detailing the following aspects: 1. The main content and theme of the video.2. The color, shape, size, texture, quantity, text, and spatial relationships of the objects.3. Actions, events, behaviors temporal relationships, physical movement changes of the objects.4. background environment, light, style and atmosphere.5. camera angles, movements, and transitions used in the video:<|eot_id|><|start_header_id|>user<|end_header_id|>\n\n{}<|eot_id|>', 'crop_start': 95}</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">max_sequence_length<span class="opacity-60">: int = 256</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">sampling_type<span class="opacity-60">: FramepackSamplingType = <FramepackSamplingType.INVERTED_ANTI_DRIFTING: 'inverted_anti_drifting'></span></span></span><!--]--> <span>)</span> <!--[0--><span class="font-bold">→</span> <span class="rounded hover:bg-gray-400 cursor-pointer"><!----><span><code>~HunyuanVideoFramepackPipelineOutput</code> or <code>tuple</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">参数 <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.HunyuanVideoFramepackPipeline.__call__.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.HunyuanVideoFramepackPipeline.__call__.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> 或 <code>np.ndarray</code> 或 <code>torch.Tensor</code>) — | |
| 用于视频生成的起始图像。<!----></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.HunyuanVideoFramepackPipeline.__call__.last_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.HunyuanVideoFramepackPipeline.__call__.last_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>last_image</strong> (<code>PIL.Image.Image</code> 或 <code>np.ndarray</code> 或 <code>torch.Tensor</code>，<em>可选</em>) — | |
| 用于视频生成的可选最后一帧图像，作为视频的结束点。这在生成两张图像之间的过渡效果时非常有用。<!----></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.HunyuanVideoFramepackPipeline.__call__.prompt" 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.HunyuanVideoFramepackPipeline.__call__.prompt"><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>prompt</strong> (<code>str</code> 或 <code>list[str]</code>，<em>可选</em>) — | |
| 用于引导图像生成的提示词或提示词列表。如果未定义，则必须改为传入 <code>prompt_embeds</code>。<!----></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.HunyuanVideoFramepackPipeline.__call__.prompt_2" 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.HunyuanVideoFramepackPipeline.__call__.prompt_2"><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>prompt_2</strong> (<code>str</code> 或 <code>list[str]</code>，<em>可选</em>) — | |
| 要发送给 <code>tokenizer_2</code> 和 <code>text_encoder_2</code> 的提示词。如果未定义，则使用 <code>prompt</code>。<!----></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.HunyuanVideoFramepackPipeline.__call__.negative_prompt" 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.HunyuanVideoFramepackPipeline.__call__.negative_prompt"><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>negative_prompt</strong> (<code>str</code> 或 <code>list[str]</code>，<em>可选</em>) — | |
| 用于不引导图像生成的提示词。如果未定义，则必须传入 <code>negative_prompt_embeds</code>。当不使用引导时会被忽略（即当 <code>true_cfg_scale</code> 不大于 <code>1</code> 时会被忽略）。<!----></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.HunyuanVideoFramepackPipeline.__call__.negative_prompt_2" 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.HunyuanVideoFramepackPipeline.__call__.negative_prompt_2"><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>negative_prompt_2</strong> (<code>str</code> 或 <code>list[str]</code>，<em>可选</em>) — | |
| 用于引导图像生成的负面提示词，将发送给 <code>tokenizer_2</code> 和 <code>text_encoder_2</code>。如果未定义，则所有文本编码器都会使用 <code>negative_prompt</code>。<!----></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.HunyuanVideoFramepackPipeline.__call__.height" 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.HunyuanVideoFramepackPipeline.__call__.height"><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>height</strong> (<code>int</code>，默认为 <code>720</code>) — | |
| 生成图像的高度（以像素为单位）。<!----></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.HunyuanVideoFramepackPipeline.__call__.width" 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.HunyuanVideoFramepackPipeline.__call__.width"><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>width</strong> (<code>int</code>，默认为 <code>1280</code>) — | |
| 生成图像的宽度（以像素为单位）。<!----></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.HunyuanVideoFramepackPipeline.__call__.num_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.HunyuanVideoFramepackPipeline.__call__.num_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>num_frames</strong> (<code>int</code>，默认为 <code>129</code>) — | |
| 生成视频的帧数。<!----></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.HunyuanVideoFramepackPipeline.__call__.latent_window_size" 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.HunyuanVideoFramepackPipeline.__call__.latent_window_size"><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>latent_window_size</strong> (<code>int</code>，默认为 <code>9</code>) — | |
| 每个 Framepack 采样窗口生成的潜在帧数量。<!----></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.HunyuanVideoFramepackPipeline.__call__.num_inference_steps" 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.HunyuanVideoFramepackPipeline.__call__.num_inference_steps"><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>num_inference_steps</strong> (<code>int</code>，默认为 <code>50</code>) — | |
| 去噪步数。更多的去噪步数通常能生成更高质量的图像，但推理速度会变慢。<!----></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.HunyuanVideoFramepackPipeline.__call__.sigmas" 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.HunyuanVideoFramepackPipeline.__call__.sigmas"><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>sigmas</strong> (<code>list[float]</code>，<em>可选</em>) — | |
| 用于在支持 <code>sigmas</code> 参数的调度器的 <code>set_timesteps</code> 方法中，为去噪过程指定自定义的 sigmas。如果未定义，则使用传入 <code>num_inference_steps</code> 时的默认行为。<!----></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.HunyuanVideoFramepackPipeline.__call__.true_cfg_scale" 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.HunyuanVideoFramepackPipeline.__call__.true_cfg_scale"><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>true_cfg_scale</strong> (<code>float</code>，<em>可选</em>，默认值为 1.0) — | |
| 当值大于 1.0 且提供了 <code>negative_prompt</code> 时，启用真正的无分类器引导（true classifier-free guidance）。<!----></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.HunyuanVideoFramepackPipeline.__call__.guidance_scale" 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.HunyuanVideoFramepackPipeline.__call__.guidance_scale"><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>guidance_scale</strong> (<code>float</code>，默认为 <code>6.0</code>) — | |
| <a href="https://huggingface.co/papers/2207.12598" rel="nofollow">Classifier-Free Diffusion | |
| Guidance</a> 中定义的引导尺度。<code>guidance_scale</code> 定义为 <a href="https://huggingface.co/papers/2205.11487" rel="nofollow">Imagen Paper</a> 中公式 2 的 <code>w</code>。通过设置 | |
| <code>guidance_scale > 1</code> 来启用引导尺度。较高的引导尺度会促使生成与文本 <code>prompt</code> 紧密关联的图像，通常以牺牲图像质量为代价。请注意，目前唯一可用的 | |
| HunyuanVideo 模型是经过 CFG 蒸馏的，这意味着不会应用无条件和条件潜在变量之间的传统引导。<!----></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.HunyuanVideoFramepackPipeline.__call__.num_videos_per_prompt" 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.HunyuanVideoFramepackPipeline.__call__.num_videos_per_prompt"><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>num_videos_per_prompt</strong> (<code>int</code>，<em>可选</em>，默认为 1) — | |
| 每个提示词生成的图像数量。<!----></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.HunyuanVideoFramepackPipeline.__call__.generator" 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.HunyuanVideoFramepackPipeline.__call__.generator"><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>generator</strong> (<code>torch.Generator</code> 或 <code>list[torch.Generator]</code>，<em>可选</em>) — | |
| 用于使生成过程确定性的 <a href="https://pytorch.org/docs/stable/generated/torch.Generator.html" rel="nofollow"><code>torch.Generator</code></a>。<!----></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.HunyuanVideoFramepackPipeline.__call__.image_latents" 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.HunyuanVideoFramepackPipeline.__call__.image_latents"><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_latents</strong> (<code>torch.Tensor</code>, <em>可选</em>) — | |
| 预编码的图像潜变量。如果未提供，将使用 VAE 对图像进行编码。<!----></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.HunyuanVideoFramepackPipeline.__call__.last_image_latents" 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.HunyuanVideoFramepackPipeline.__call__.last_image_latents"><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>last_image_latents</strong> (<code>torch.Tensor</code>, <em>可选</em>) — | |
| 预编码的最后一帧图像潜变量。如果未提供，将使用 VAE 对最后一帧图像进行编码。<!----></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.HunyuanVideoFramepackPipeline.__call__.prompt_embeds" 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.HunyuanVideoFramepackPipeline.__call__.prompt_embeds"><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>prompt_embeds</strong> (<code>torch.Tensor</code>, <em>可选</em>) — | |
| 预生成的文本嵌入。可用于轻松调整文本输入（提示词加权）。如果未提供，则从 <code>prompt</code> 输入参数生成文本嵌入。<!----></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.HunyuanVideoFramepackPipeline.__call__.pooled_prompt_embeds" 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.HunyuanVideoFramepackPipeline.__call__.pooled_prompt_embeds"><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>pooled_prompt_embeds</strong> (<code>torch.FloatTensor</code>，<em>可选</em>) — | |
| 预生成的池化文本嵌入。可用于轻松调整文本输入，<em>例如</em>提示词加权。 | |
| 如果未提供，将从 <code>prompt</code> 输入参数生成池化文本嵌入。<!----></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.HunyuanVideoFramepackPipeline.__call__.negative_prompt_embeds" 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.HunyuanVideoFramepackPipeline.__call__.negative_prompt_embeds"><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>negative_prompt_embeds</strong> (<code>torch.FloatTensor</code>, <em>可选</em>) — | |
| 预生成的负向文本嵌入。可用于轻松调整文本输入，<em>例如</em>提示词加权。如果未提供，negative_prompt_embeds 将根据 <code>negative_prompt</code> 输入参数生成。<!----></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.HunyuanVideoFramepackPipeline.__call__.negative_pooled_prompt_embeds" 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.HunyuanVideoFramepackPipeline.__call__.negative_pooled_prompt_embeds"><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>negative_pooled_prompt_embeds</strong> (<code>torch.FloatTensor</code>，<em>可选</em>) — | |
| 预生成的负向池化文本嵌入。可用于轻松调整文本输入，<em>例如</em>提示词加权。如果未提供，将从 <code>negative_prompt</code> 输入参数生成池化的 negative_prompt_embeds。<!----></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.HunyuanVideoFramepackPipeline.__call__.prompt_attention_mask" 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.HunyuanVideoFramepackPipeline.__call__.prompt_attention_mask"><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>prompt_attention_mask</strong> (<code>torch.Tensor</code>, <em>可选</em>) — | |
| <code>prompt_embeds</code> 的注意力掩码。当直接传入 <code>prompt_embeds</code> 时，此参数为必填项。<!----></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.HunyuanVideoFramepackPipeline.__call__.negative_prompt_attention_mask" 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.HunyuanVideoFramepackPipeline.__call__.negative_prompt_attention_mask"><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>negative_prompt_attention_mask</strong> (<code>torch.Tensor</code>, <em>可选</em>) — | |
| <code>negative_prompt_embeds</code> 的注意力掩码。当直接传入 <code>negative_prompt_embeds</code> 时，此参数为必填项。<!----></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.HunyuanVideoFramepackPipeline.__call__.output_type" 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.HunyuanVideoFramepackPipeline.__call__.output_type"><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>output_type</strong> (<code>str</code>，<em>可选</em>，默认为 <code>"pil"</code>) — | |
| 生成图像的输出格式。可选 <code>PIL.Image</code> 或 <code>np.array</code>。<!----></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.HunyuanVideoFramepackPipeline.__call__.return_dict" 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.HunyuanVideoFramepackPipeline.__call__.return_dict"><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>return_dict</strong> (<code>bool</code>，<em>可选</em>，默认为 <code>True</code>) — | |
| 是否返回 <code>HunyuanVideoFramepackPipelineOutput</code> 而不是普通元组。<!----></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.HunyuanVideoFramepackPipeline.__call__.attention_kwargs" 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.HunyuanVideoFramepackPipeline.__call__.attention_kwargs"><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>attention_kwargs</strong> (<code>dict</code>, <em>可选</em>) — | |
| 一个 kwargs 字典，如果指定，将传递给 <a href="https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py" rel="nofollow">diffusers.models.attention_processor</a> 中定义的 <code>AttentionProcessor</code>，该处理器在 <code>self.processor</code> 下定义。<!----></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.HunyuanVideoFramepackPipeline.__call__.prompt_template" 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.HunyuanVideoFramepackPipeline.__call__.prompt_template"><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>prompt_template</strong> (<code>dict</code>, <em>可选</em>) — | |
| 用于在编码前格式化提示词的模板。默认使用模型的默认模板。<!----></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.HunyuanVideoFramepackPipeline.__call__.max_sequence_length" 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.HunyuanVideoFramepackPipeline.__call__.max_sequence_length"><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>max_sequence_length</strong> (<code>int</code>，<em>可选</em>，默认为 256) — | |
| 用于提示编码器的最大序列长度。<!----></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.HunyuanVideoFramepackPipeline.__call__.sampling_type" 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.HunyuanVideoFramepackPipeline.__call__.sampling_type"><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>sampling_type</strong> (<code>FramepackSamplingType</code>，<em>可选</em>) — | |
| 在遍历潜在窗口时使用的 Framepack 采样策略。<!----></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.HunyuanVideoFramepackPipeline.__call__.callback_on_step_end" 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.HunyuanVideoFramepackPipeline.__call__.callback_on_step_end"><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>callback_on_step_end</strong> (<code>Callable</code>、<code>PipelineCallback</code>、<code>MultiPipelineCallbacks</code>、<em>可选</em>) — | |
| 在推理过程中，每个去噪步骤结束时调用的函数，或 <code>PipelineCallback</code> 或 <code>MultiPipelineCallbacks</code> 的子类，其参数如下：<code>callback_on_step_end(self: DiffusionPipeline, step: int, timestep: int, callback_kwargs: Dict)</code>。<code>callback_kwargs</code> 将包含由 <code>callback_on_step_end_tensor_inputs</code> 指定的所有张量列表。<!----></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.HunyuanVideoFramepackPipeline.__call__.callback_on_step_end_tensor_inputs" 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.HunyuanVideoFramepackPipeline.__call__.callback_on_step_end_tensor_inputs"><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>callback_on_step_end_tensor_inputs</strong> (<code>list</code>，<em>可选</em>) — | |
| <code>callback_on_step_end</code> 函数的张量输入列表。列表中指定的张量将作为 <code>callback_kwargs</code> 参数传递。您只能包含流水线类 <code>._callback_tensor_inputs</code> 属性中列出的变量。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[0--><div id="diffusers.HunyuanVideoFramepackPipeline.__call__.returns" class="flex items-center font-semibold space-x-3 text-base !mt-0 !mb-0 text-gray-800 rounded "><p class="text-base"><!---->返回值<!----></p> <!--[0--><!----> | |
| <p><code>~HunyuanVideoFramepackPipelineOutput</code> 或 <code>tuple</code></p> | |
| <!----><!--]--> <span class="flex-auto border-t-2 border-gray-100 dark:border-gray-700"></span></div> <div class="text-base"><!----> | |
| <p>如果 <code>return_dict</code> 为 <code>True</code>，则返回 <code>HunyuanVideoFramepackPipelineOutput</code>；否则返回一个 <code>tuple</code>，其中第一个元素是包含生成图像的列表，第二个元素是 <code>bool</code> 列表，用于指示对应的生成图像是否包含“不适合工作场所”（nsfw）内容。</p> | |
| <!----></div><!--]--> <!--[-1--><!--]--></div></div><!----> <p>用于生成的 pipeline 的调用函数。</p> <p>示例：</p> <!--[4--><h5 class="relative group"><a id="image-to-video" 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="#image-to-video"><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><!---->图像到视频<!----></span></h5><!--]--><!----> <div class="relative group rounded-md"><a id="diffusers.HunyuanVideoFramepackPipeline.__call__.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.HunyuanVideoFramepackPipeline.__call__.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> 已复制</div><!----></button><!----></div> <pre class="language-python "><!----><span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> HunyuanVideoFramepackPipeline, HunyuanVideoFramepackTransformer3DModel | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> export_to_video, load_image | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> SiglipImageProcessor, SiglipVisionModel | |
| <span class="hljs-meta">>>> </span>transformer = HunyuanVideoFramepackTransformer3DModel.from_pretrained( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"lllyasviel/FramePackI2V_HY"</span>, torch_dtype=torch.bfloat16 | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>feature_extractor = SiglipImageProcessor.from_pretrained( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"feature_extractor"</span> | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>image_encoder = SiglipVisionModel.from_pretrained( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"image_encoder"</span>, torch_dtype=torch.float16 | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>pipe = HunyuanVideoFramepackPipeline.from_pretrained( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"hunyuanvideo-community/HunyuanVideo"</span>, | |
| <span class="hljs-meta">... </span> transformer=transformer, | |
| <span class="hljs-meta">... </span> feature_extractor=feature_extractor, | |
| <span class="hljs-meta">... </span> image_encoder=image_encoder, | |
| <span class="hljs-meta">... </span> torch_dtype=torch.float16, | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>pipe.vae.enable_tiling() | |
| <span class="hljs-meta">>>> </span>pipe.to(<span class="hljs-string">"cuda"</span>) | |
| <span class="hljs-meta">>>> </span>image = load_image( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/penguin.png"</span> | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>output = pipe( | |
| <span class="hljs-meta">... </span> image=image, | |
| <span class="hljs-meta">... </span> prompt=<span class="hljs-string">"A penguin dancing in the snow"</span>, | |
| <span class="hljs-meta">... </span> height=<span class="hljs-number">832</span>, | |
| <span class="hljs-meta">... </span> width=<span class="hljs-number">480</span>, | |
| <span class="hljs-meta">... </span> num_frames=<span class="hljs-number">91</span>, | |
| <span class="hljs-meta">... </span> num_inference_steps=<span class="hljs-number">30</span>, | |
| <span class="hljs-meta">... </span> guidance_scale=<span class="hljs-number">9.0</span>, | |
| <span class="hljs-meta">... </span> generator=torch.Generator().manual_seed(<span class="hljs-number">0</span>), | |
| <span class="hljs-meta">... </span> sampling_type=<span class="hljs-string">"inverted_anti_drifting"</span>, | |
| <span class="hljs-meta">... </span>).frames[<span class="hljs-number">0</span>] | |
| <span class="hljs-meta">>>> </span>export_to_video(output, <span class="hljs-string">"output.mp4"</span>, fps=<span class="hljs-number">30</span>)<!----></pre></div><!----></div><!----> <!--[4--><h5 class="relative group"><a id="first-and-last-image-to-video" 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="#first-and-last-image-to-video"><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><!---->首帧和尾帧图像转视频<!----></span></h5><!--]--><!----> <div class="relative group rounded-md"><a id="diffusers.HunyuanVideoFramepackPipeline.__call__.example-2" 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.HunyuanVideoFramepackPipeline.__call__.example-2"><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> 已复制</div><!----></button><!----></div> <pre class="language-python "><!----><span class="hljs-meta">>>> </span><span class="hljs-keyword">import</span> torch | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> diffusers <span class="hljs-keyword">import</span> HunyuanVideoFramepackPipeline, HunyuanVideoFramepackTransformer3DModel | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> diffusers.utils <span class="hljs-keyword">import</span> export_to_video, load_image | |
| <span class="hljs-meta">>>> </span><span class="hljs-keyword">from</span> transformers <span class="hljs-keyword">import</span> SiglipImageProcessor, SiglipVisionModel | |
| <span class="hljs-meta">>>> </span>transformer = HunyuanVideoFramepackTransformer3DModel.from_pretrained( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"lllyasviel/FramePackI2V_HY"</span>, torch_dtype=torch.bfloat16 | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>feature_extractor = SiglipImageProcessor.from_pretrained( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"feature_extractor"</span> | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>image_encoder = SiglipVisionModel.from_pretrained( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"lllyasviel/flux_redux_bfl"</span>, subfolder=<span class="hljs-string">"image_encoder"</span>, torch_dtype=torch.float16 | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>pipe = HunyuanVideoFramepackPipeline.from_pretrained( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"hunyuanvideo-community/HunyuanVideo"</span>, | |
| <span class="hljs-meta">... </span> transformer=transformer, | |
| <span class="hljs-meta">... </span> feature_extractor=feature_extractor, | |
| <span class="hljs-meta">... </span> image_encoder=image_encoder, | |
| <span class="hljs-meta">... </span> torch_dtype=torch.float16, | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>pipe.to(<span class="hljs-string">"cuda"</span>) | |
| <span class="hljs-meta">>>> </span>prompt = <span class="hljs-string">"CG animation style, a small blue bird takes off from the ground, flapping its wings. The bird's feathers are delicate, with a unique pattern on its chest. The background shows a blue sky with white clouds under bright sunshine. The camera follows the bird upward, capturing its flight and the vastness of the sky from a close-up, low-angle perspective."</span> | |
| <span class="hljs-meta">>>> </span>first_image = load_image( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/flf2v_input_first_frame.png"</span> | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>last_image = load_image( | |
| <span class="hljs-meta">... </span> <span class="hljs-string">"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/flf2v_input_last_frame.png"</span> | |
| <span class="hljs-meta">... </span>) | |
| <span class="hljs-meta">>>> </span>output = pipe( | |
| <span class="hljs-meta">... </span> image=first_image, | |
| <span class="hljs-meta">... </span> last_image=last_image, | |
| <span class="hljs-meta">... </span> prompt=prompt, | |
| <span class="hljs-meta">... </span> height=<span class="hljs-number">512</span>, | |
| <span class="hljs-meta">... </span> width=<span class="hljs-number">512</span>, | |
| <span class="hljs-meta">... </span> num_frames=<span class="hljs-number">91</span>, | |
| <span class="hljs-meta">... </span> num_inference_steps=<span class="hljs-number">30</span>, | |
| <span class="hljs-meta">... </span> guidance_scale=<span class="hljs-number">9.0</span>, | |
| <span class="hljs-meta">... </span> generator=torch.Generator().manual_seed(<span class="hljs-number">0</span>), | |
| <span class="hljs-meta">... </span> sampling_type=<span class="hljs-string">"inverted_anti_drifting"</span>, | |
| <span class="hljs-meta">... </span>).frames[<span class="hljs-number">0</span>] | |
| <span class="hljs-meta">>>> </span>export_to_video(output, <span class="hljs-string">"output.mp4"</span>, fps=<span class="hljs-number">30</span>)<!----></pre></div><!----></div><!----></div></div> <!--[1--><h2 class="relative group"><a id="diffusers.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput" 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.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput"><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><!---->HunyuanVideoPipelineOutput<!----></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.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput"><!----><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.pipelines.hunyuan_video.pipeline_output.</span><span class="font-semibold">HunyuanVideoPipelineOutput</span></span></h3><!----> <a id="diffusers.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput" class="header-link invisible with-hover:group-hover:visible pr-2" href="#diffusers.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput"><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/v0.43.0/src/diffusers/pipelines/hunyuan_video/pipeline_output.py#L11" target="_blank"><span><</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">: 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">参数 <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.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput.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.pipelines.hunyuan_video.pipeline_output.HunyuanVideoPipelineOutput.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>torch.Tensor</code>、<code>np.ndarray</code> 或 list[list[PIL.Image.Image]]) — | |
| 视频输出列表 - 它可以是一个长度为 <code>batch_size,</code> 的嵌套列表，其中每个子列表包含长度为 <code>num_frames.</code> 的去噪 PIL 图像序列。它也可以是一个形状为 <code>(batch_size, num_frames, channels, height, width)</code> 的 NumPy 数组或 Torch 张量。<!----></span></span></li><!--]--></ul><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--> <!--[-1--><!--]--></div></div><!----> <p>HunyuanVideo 管道的输出类。</p></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/framepack.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_ywq9g6 = { | |
| base: "/docs/diffusers/v0.43.0/en", | |
| assets: "/docs/diffusers/v0.43.0/en" | |
| }; | |
| const element = document.currentScript.parentElement; | |
| Promise.all([ | |
| import("/docs/diffusers/v0.43.0/en/_app/immutable/entry/start.1gaXe03k.js"), | |
| import("/docs/diffusers/v0.43.0/en/_app/immutable/entry/app.BwdYRbCn.js") | |
| ]).then(([kit, app]) => { | |
| kit.start(app, element, { | |
| node_ids: [0, 181], | |
| data: [null,null], | |
| form: null, | |
| error: null | |
| }); | |
| }); | |
| } | |
| </script> | |
Xet Storage Details
- Size:
- 126 kB
- Xet hash:
- b955cab53e3caef52c8c57537bbfb430e1208f6c7d49a86b73f1729c4046df10
·
Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.