<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by zhangp365 on Medium]]></title>
        <description><![CDATA[Stories by zhangp365 on Medium]]></description>
        <link>https://medium.com/@zhangp365?source=rss-aff5b606c38a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*1LngEeOP_vzxg6gxADPv5Q.png</url>
            <title>Stories by zhangp365 on Medium</title>
            <link>https://medium.com/@zhangp365?source=rss-aff5b606c38a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 03:20:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@zhangp365/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Introduction to the comfyUI + animateDiff Video2Video, AI Video Generation Workflow and Examples]]></title>
            <link>https://medium.com/@zhangp365/introduction-to-the-comfyui-animatediff-video2video-ai-video-generation-workflow-and-examples-f6cec78f640b?source=rss-aff5b606c38a------2</link>
            <guid isPermaLink="false">https://medium.com/p/f6cec78f640b</guid>
            <category><![CDATA[ai-video-generation]]></category>
            <category><![CDATA[latent-consistency-model]]></category>
            <category><![CDATA[lcms]]></category>
            <category><![CDATA[animatediff]]></category>
            <category><![CDATA[comfy-ui]]></category>
            <dc:creator><![CDATA[zhangp365]]></dc:creator>
            <pubDate>Sun, 10 Dec 2023 07:33:18 GMT</pubDate>
            <atom:updated>2024-01-07T03:27:29.314Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Update:</strong></h3><p>As of January 7, 2024, the animatediff v3 model has been released. I have upgraded the previous animatediff model to the v3 version and updated the workflow accordingly, resulting in newly generated videos.</p><p>Recently, there has been a surge in popularity for creating videos using the stable diffusion + animateDiff techniques. However, for the average user, successfully generating a video on their computer can be a challenging task. This article aims to guide you through the process of setting up the workflow for loading comfyUI + animateDiff and producing related videos. The article is divided into the following key sections:</p><ol><li>Setting Up the Working Environment</li><li>Generating the first video</li><li>Progressing to generate additional videos</li><li>Introduction to Important Considerations</li></ol><h3>Setting Up the Working Environment</h3><p><strong>Introduction to comfyUI</strong></p><p><a href="https://github.com/comfyanonymous/ComfyUI">comfyUI</a> stands out as an AI drawing software with a versatile node-based and flow-style custom workflow. It offers convenient functionalities such as text-to-image, graphic generation, image upscaling, inpainting, and the loading of controlnet control for generation. Moreover, as demonstrated in the workflows provided later in this article, comfyUI is a superior choice for video generation compared to other AI drawing software, offering higher efficiency and better results.</p><p><strong>Installing comfyUI</strong></p><p>For detailed instructions, please refer to the <a href="https://github.com/comfyanonymous/ComfyUI">comfyUI</a> page. After installing the Python environment, follow the step-by-step process to install the necessary dependencies, ultimately completing the installation of comfyUI. Familiarity with the Python language and pip installation is required for this process. This article does not provide a detailed installation guide, but you can search and find solutions if needed. For Windows users, an alternative option is to download a pre-packaged Python environment directly from the comfyUI project introduction. After extracting the package, you can try running it. If comfyUI is already installed on your computer, you can proceed to the next step of loading workflows and videos to start the generation.</p><p>For regular users who prefer not to manually install the Python environment and related software, you can download and install <a href="https://www.tensorbee.com/">Tensorbee</a> from the official Tensorbee website. Alternatively, you can directly click on <a href="https://api.tensorbee.com/api/downloadtensorbee?channel=tb_0_3_0">the download link</a> for Tensorbee. After installing tensorbee, within tensorbee, click to download AnimateComfy. Tensorbee will then configure the comfyUI working environment and the workflow used in this article. You only need to click “generate” to create your first video.</p><h3>Generating the first video</h3><p>The initial video for the first video generation is:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FSIu8fJCd0iA%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DSIu8fJCd0iA&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FSIu8fJCd0iA%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/b7e1b53aa5772653b1b2491c05924589/href">https://medium.com/media/b7e1b53aa5772653b1b2491c05924589/href</a></iframe><p>Our objective is to have AI learn the hand gestures and actions in this video, ultimately producing a new video.</p><p>To accomplish this, we will utilize the following workflow:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*E6Pu_C3k_ViWXGcBhlzuSA.png" /></figure><p>Note: If you are a Tensorbee user, simply click the “Queue Prompt” on the right to start the generation. You can skip the section on loading workflows and related content.</p><p>You can download the following JSON and save it to the .json file, then use the “Load” button on the right side of comfyUI to load the file and the workflow mentioned above.</p><pre>{&quot;last_node_id&quot;:116,&quot;last_link_id&quot;:277,&quot;nodes&quot;:[{&quot;id&quot;:92,&quot;type&quot;:&quot;PreviewImage&quot;,&quot;pos&quot;:[1740.2310688330463,73.08600474074845],&quot;size&quot;:{&quot;0&quot;:210,&quot;1&quot;:246},&quot;flags&quot;:{},&quot;order&quot;:17,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;images&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;link&quot;:220,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;images&quot;}],&quot;title&quot;:&quot;controlNet中间预览&quot;,&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;PreviewImage&quot;,&quot;ttNbgOverride&quot;:{&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;,&quot;groupcolor&quot;:&quot;#88A&quot;}},&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;},{&quot;id&quot;:5,&quot;type&quot;:&quot;PrimitiveNode&quot;,&quot;pos&quot;:[108.43046940534282,567.1134323568291],&quot;size&quot;:{&quot;0&quot;:210,&quot;1&quot;:82},&quot;flags&quot;:{},&quot;order&quot;:0,&quot;mode&quot;:0,&quot;outputs&quot;:[{&quot;name&quot;:&quot;INT&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;links&quot;:[237],&quot;slot_index&quot;:0,&quot;widget&quot;:{&quot;name&quot;:&quot;seed&quot;},&quot;label&quot;:&quot;INT&quot;}],&quot;title&quot;:&quot;seed 种子值&quot;,&quot;properties&quot;:{&quot;Run widget replace on values&quot;:false},&quot;widgets_values&quot;:[34444447,&quot;fixed&quot;],&quot;color&quot;:&quot;#2a363b&quot;,&quot;bgcolor&quot;:&quot;#3f5159&quot;},{&quot;id&quot;:70,&quot;type&quot;:&quot;ControlNetLoaderAdvanced&quot;,&quot;pos&quot;:[1335,250],&quot;size&quot;:{&quot;0&quot;:367.79998779296877,&quot;1&quot;:58},&quot;flags&quot;:{},&quot;order&quot;:1,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;timestep_keyframe&quot;,&quot;type&quot;:&quot;TIMESTEP_KEYFRAME&quot;,&quot;link&quot;:null,&quot;label&quot;:&quot;timestep_keyframe&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;CONTROL_NET&quot;,&quot;type&quot;:&quot;CONTROL_NET&quot;,&quot;links&quot;:[118],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;CONTROL_NET&quot;}],&quot;title&quot;:&quot;Load ControlNet Model 模型加载器&quot;,&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;ControlNetLoaderAdvanced&quot;},&quot;widgets_values&quot;:[&quot;control_v11p_sd15_openpose.pth&quot;],&quot;color&quot;:&quot;#2a363b&quot;,&quot;bgcolor&quot;:&quot;#3f5159&quot;},{&quot;id&quot;:10,&quot;type&quot;:&quot;VAEDecode&quot;,&quot;pos&quot;:[1732,776],&quot;size&quot;:{&quot;0&quot;:210,&quot;1&quot;:46},&quot;flags&quot;:{},&quot;order&quot;:22,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;samples&quot;,&quot;type&quot;:&quot;LATENT&quot;,&quot;link&quot;:245,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;samples&quot;},{&quot;name&quot;:&quot;vae&quot;,&quot;type&quot;:&quot;VAE&quot;,&quot;link&quot;:275,&quot;label&quot;:&quot;vae&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;IMAGE&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;links&quot;:[130,221],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;IMAGE&quot;}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;VAEDecode&quot;,&quot;ttNbgOverride&quot;:{&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;,&quot;groupcolor&quot;:&quot;#88A&quot;}},&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;},{&quot;id&quot;:103,&quot;type&quot;:&quot;VHS_VideoCombine&quot;,&quot;pos&quot;:[2159,952],&quot;size&quot;:[369.5643310546875,432.376220703125],&quot;flags&quot;:{},&quot;order&quot;:24,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;images&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;link&quot;:221,&quot;label&quot;:&quot;images&quot;}],&quot;outputs&quot;:[],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;VHS_VideoCombine&quot;},&quot;widgets_values&quot;:{&quot;frame_rate&quot;:12,&quot;loop_count&quot;:0,&quot;filename_prefix&quot;:&quot;AnimateDiff&quot;,&quot;format&quot;:&quot;video/h265-mp4&quot;,&quot;pingpong&quot;:false,&quot;save_image&quot;:true,&quot;crf&quot;:20,&quot;videopreview&quot;:{&quot;hidden&quot;:false,&quot;paused&quot;:false,&quot;params&quot;:{&quot;filename&quot;:&quot;AnimateDiff_00331.mp4&quot;,&quot;subfolder&quot;:&quot;&quot;,&quot;type&quot;:&quot;output&quot;,&quot;format&quot;:&quot;video/h265-mp4&quot;}}}},{&quot;id&quot;:93,&quot;type&quot;:&quot;ADE_AnimateDiffLoaderWithContext&quot;,&quot;pos&quot;:[1483,1491],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:190},&quot;flags&quot;:{},&quot;order&quot;:20,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;model&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;link&quot;:269,&quot;label&quot;:&quot;model&quot;},{&quot;name&quot;:&quot;context_options&quot;,&quot;type&quot;:&quot;CONTEXT_OPTIONS&quot;,&quot;link&quot;:206,&quot;slot_index&quot;:1,&quot;label&quot;:&quot;context_options&quot;},{&quot;name&quot;:&quot;motion_lora&quot;,&quot;type&quot;:&quot;MOTION_LORA&quot;,&quot;link&quot;:null,&quot;label&quot;:&quot;motion_lora&quot;},{&quot;name&quot;:&quot;motion_model_settings&quot;,&quot;type&quot;:&quot;MOTION_MODEL_SETTINGS&quot;,&quot;link&quot;:null,&quot;label&quot;:&quot;motion_model_settings&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;MODEL&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;links&quot;:[241],&quot;shape&quot;:3,&quot;label&quot;:&quot;MODEL&quot;}],&quot;title&quot;:&quot;AnimateDiff Loader动画模型加载器&quot;,&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;ADE_AnimateDiffLoaderWithContext&quot;,&quot;ttNbgOverride&quot;:{&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;,&quot;groupcolor&quot;:&quot;#88A&quot;}},&quot;widgets_values&quot;:[&quot;v3_sd15_mm.ckpt&quot;,&quot;sqrt_linear (AnimateDiff)&quot;,1,false],&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;},{&quot;id&quot;:94,&quot;type&quot;:&quot;ADE_AnimateDiffUniformContextOptions&quot;,&quot;pos&quot;:[1100,1273],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:154},&quot;flags&quot;:{},&quot;order&quot;:2,&quot;mode&quot;:0,&quot;outputs&quot;:[{&quot;name&quot;:&quot;CONTEXT_OPTIONS&quot;,&quot;type&quot;:&quot;CONTEXT_OPTIONS&quot;,&quot;links&quot;:[206],&quot;shape&quot;:3,&quot;label&quot;:&quot;CONTEXT_OPTIONS&quot;}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;ADE_AnimateDiffUniformContextOptions&quot;,&quot;ttNbgOverride&quot;:{&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;,&quot;groupcolor&quot;:&quot;#88A&quot;}},&quot;widgets_values&quot;:[16,1,4,&quot;uniform&quot;,false],&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;},{&quot;id&quot;:100,&quot;type&quot;:&quot;VHS_LoadVideo&quot;,&quot;pos&quot;:[86.4304694053429,103.11343235682848],&quot;size&quot;:[315,386.3125],&quot;flags&quot;:{},&quot;order&quot;:3,&quot;mode&quot;:0,&quot;outputs&quot;:[{&quot;name&quot;:&quot;IMAGE&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;links&quot;:[262],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;IMAGE&quot;},{&quot;name&quot;:&quot;frame_count&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;links&quot;:[259],&quot;shape&quot;:3,&quot;slot_index&quot;:1,&quot;label&quot;:&quot;frame_count&quot;}],&quot;title&quot;:&quot;Load Video 视频加载器&quot;,&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;VHS_LoadVideo&quot;,&quot;ttNbgOverride&quot;:{&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;,&quot;groupcolor&quot;:&quot;#88A&quot;}},&quot;widgets_values&quot;:{&quot;video&quot;:&quot;example.mp4&quot;,&quot;force_rate&quot;:0,&quot;force_size&quot;:&quot;Disabled&quot;,&quot;custom_width&quot;:512,&quot;custom_height&quot;:512,&quot;frame_load_cap&quot;:1200,&quot;skip_first_frames&quot;:0,&quot;select_every_nth&quot;:2,&quot;choose video to upload&quot;:&quot;image&quot;,&quot;videopreview&quot;:{&quot;hidden&quot;:false,&quot;paused&quot;:false,&quot;params&quot;:{&quot;filename&quot;:&quot;example.mp4&quot;,&quot;type&quot;:&quot;input&quot;,&quot;format&quot;:&quot;video&quot;}}},&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;},{&quot;id&quot;:97,&quot;type&quot;:&quot;PrimitiveNode&quot;,&quot;pos&quot;:[15.430469405342637,706.1134323568294],&quot;size&quot;:{&quot;0&quot;:210,&quot;1&quot;:82},&quot;flags&quot;:{},&quot;order&quot;:4,&quot;mode&quot;:0,&quot;outputs&quot;:[{&quot;name&quot;:&quot;INT&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;links&quot;:[260,263],&quot;slot_index&quot;:0,&quot;widget&quot;:{&quot;name&quot;:&quot;width&quot;},&quot;label&quot;:&quot;INT&quot;}],&quot;title&quot;:&quot;Width宽度&quot;,&quot;properties&quot;:{&quot;Run widget replace on values&quot;:false},&quot;widgets_values&quot;:[768,&quot;fixed&quot;],&quot;color&quot;:&quot;#332922&quot;,&quot;bgcolor&quot;:&quot;#593930&quot;},{&quot;id&quot;:112,&quot;type&quot;:&quot;EmptyLatentImage&quot;,&quot;pos&quot;:[747,573],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:106},&quot;flags&quot;:{},&quot;order&quot;:8,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;width&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;link&quot;:260,&quot;widget&quot;:{&quot;name&quot;:&quot;width&quot;},&quot;label&quot;:&quot;width&quot;},{&quot;name&quot;:&quot;height&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;link&quot;:257,&quot;widget&quot;:{&quot;name&quot;:&quot;height&quot;},&quot;label&quot;:&quot;height&quot;},{&quot;name&quot;:&quot;batch_size&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;link&quot;:259,&quot;widget&quot;:{&quot;name&quot;:&quot;batch_size&quot;},&quot;label&quot;:&quot;batch_size&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;LATENT&quot;,&quot;type&quot;:&quot;LATENT&quot;,&quot;links&quot;:[261],&quot;shape&quot;:3,&quot;label&quot;:&quot;LATENT&quot;,&quot;slot_index&quot;:0}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;EmptyLatentImage&quot;},&quot;widgets_values&quot;:[768,512,1]},{&quot;id&quot;:98,&quot;type&quot;:&quot;PrimitiveNode&quot;,&quot;pos&quot;:[276,707],&quot;size&quot;:{&quot;0&quot;:210,&quot;1&quot;:82},&quot;flags&quot;:{},&quot;order&quot;:5,&quot;mode&quot;:0,&quot;outputs&quot;:[{&quot;name&quot;:&quot;INT&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;links&quot;:[257,267],&quot;slot_index&quot;:0,&quot;widget&quot;:{&quot;name&quot;:&quot;height&quot;},&quot;label&quot;:&quot;INT&quot;}],&quot;title&quot;:&quot;Height高度&quot;,&quot;properties&quot;:{&quot;Run widget replace on values&quot;:false},&quot;widgets_values&quot;:[512,&quot;fixed&quot;],&quot;color&quot;:&quot;#332922&quot;,&quot;bgcolor&quot;:&quot;#593930&quot;},{&quot;id&quot;:105,&quot;type&quot;:&quot;ModelSamplingDiscrete&quot;,&quot;pos&quot;:[733,1493],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:82},&quot;flags&quot;:{},&quot;order&quot;:16,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;model&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;link&quot;:225,&quot;label&quot;:&quot;model&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;MODEL&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;links&quot;:[268],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;MODEL&quot;}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;ModelSamplingDiscrete&quot;},&quot;widgets_values&quot;:[&quot;lcm&quot;,false]},{&quot;id&quot;:104,&quot;type&quot;:&quot;LoraLoader&quot;,&quot;pos&quot;:[729,1299],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:126},&quot;flags&quot;:{},&quot;order&quot;:11,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;model&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;link&quot;:274,&quot;label&quot;:&quot;model&quot;},{&quot;name&quot;:&quot;clip&quot;,&quot;type&quot;:&quot;CLIP&quot;,&quot;link&quot;:277,&quot;label&quot;:&quot;clip&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;MODEL&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;links&quot;:[225],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;MODEL&quot;},{&quot;name&quot;:&quot;CLIP&quot;,&quot;type&quot;:&quot;CLIP&quot;,&quot;links&quot;:[],&quot;shape&quot;:3,&quot;slot_index&quot;:1,&quot;label&quot;:&quot;CLIP&quot;}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;LoraLoader&quot;},&quot;widgets_values&quot;:[&quot;pytorch_lora_weights_lcm_1.5.safetensors&quot;,1,1]},{&quot;id&quot;:114,&quot;type&quot;:&quot;LoraLoaderModelOnly&quot;,&quot;pos&quot;:[732,1645],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:82},&quot;flags&quot;:{},&quot;order&quot;:19,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;model&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;link&quot;:268,&quot;label&quot;:&quot;model&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;MODEL&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;links&quot;:[269],&quot;shape&quot;:3,&quot;label&quot;:&quot;MODEL&quot;,&quot;slot_index&quot;:0}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;LoraLoaderModelOnly&quot;},&quot;widgets_values&quot;:[&quot;v3_sd15_adapter.ckpt&quot;,1]},{&quot;id&quot;:113,&quot;type&quot;:&quot;ImageResize+&quot;,&quot;pos&quot;:[739,75],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:170},&quot;flags&quot;:{},&quot;order&quot;:9,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;image&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;link&quot;:262,&quot;label&quot;:&quot;image&quot;},{&quot;name&quot;:&quot;width&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;link&quot;:263,&quot;widget&quot;:{&quot;name&quot;:&quot;width&quot;},&quot;label&quot;:&quot;width&quot;},{&quot;name&quot;:&quot;height&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;link&quot;:267,&quot;widget&quot;:{&quot;name&quot;:&quot;height&quot;},&quot;label&quot;:&quot;height&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;IMAGE&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;links&quot;:[265,266],&quot;shape&quot;:3,&quot;label&quot;:&quot;IMAGE&quot;,&quot;slot_index&quot;:0},{&quot;name&quot;:&quot;width&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;links&quot;:null,&quot;shape&quot;:3,&quot;label&quot;:&quot;width&quot;},{&quot;name&quot;:&quot;height&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;links&quot;:null,&quot;shape&quot;:3,&quot;label&quot;:&quot;height&quot;}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;ImageResize+&quot;},&quot;widgets_values&quot;:[768,512,&quot;nearest&quot;,true]},{&quot;id&quot;:102,&quot;type&quot;:&quot;DWPreprocessor&quot;,&quot;pos&quot;:[739,322],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:178},&quot;flags&quot;:{},&quot;order&quot;:13,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;image&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;link&quot;:266,&quot;label&quot;:&quot;image&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;IMAGE&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;links&quot;:[220,253],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;IMAGE&quot;}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;DWPreprocessor&quot;},&quot;widgets_values&quot;:[&quot;enable&quot;,&quot;enable&quot;,&quot;disable&quot;,512,&quot;yolox_l.onnx&quot;,&quot;dw-ll_ucoco_384.onnx&quot;]},{&quot;id&quot;:50,&quot;type&quot;:&quot;PreviewImage&quot;,&quot;pos&quot;:[1096,72],&quot;size&quot;:{&quot;0&quot;:210,&quot;1&quot;:246},&quot;flags&quot;:{},&quot;order&quot;:12,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;images&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;link&quot;:265,&quot;label&quot;:&quot;images&quot;}],&quot;title&quot;:&quot;Preview Image图片预览&quot;,&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;PreviewImage&quot;,&quot;ttNbgOverride&quot;:{&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;,&quot;groupcolor&quot;:&quot;#88A&quot;}},&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;},{&quot;id&quot;:108,&quot;type&quot;:&quot;KSampler&quot;,&quot;pos&quot;:[1477.231068833049,927.0860047407488],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:262},&quot;flags&quot;:{},&quot;order&quot;:21,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;model&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;link&quot;:241,&quot;label&quot;:&quot;model&quot;},{&quot;name&quot;:&quot;positive&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;link&quot;:240,&quot;label&quot;:&quot;positive&quot;},{&quot;name&quot;:&quot;negative&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;link&quot;:239,&quot;label&quot;:&quot;negative&quot;},{&quot;name&quot;:&quot;latent_image&quot;,&quot;type&quot;:&quot;LATENT&quot;,&quot;link&quot;:261,&quot;label&quot;:&quot;latent_image&quot;},{&quot;name&quot;:&quot;seed&quot;,&quot;type&quot;:&quot;INT&quot;,&quot;link&quot;:237,&quot;widget&quot;:{&quot;name&quot;:&quot;seed&quot;},&quot;label&quot;:&quot;seed&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;LATENT&quot;,&quot;type&quot;:&quot;LATENT&quot;,&quot;links&quot;:[245],&quot;shape&quot;:3,&quot;label&quot;:&quot;LATENT&quot;}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;KSampler&quot;},&quot;widgets_values&quot;:[34444447,&quot;randomize&quot;,6,1,&quot;lcm&quot;,&quot;sgm_uniform&quot;,1]},{&quot;id&quot;:72,&quot;type&quot;:&quot;ControlNetApplyAdvanced&quot;,&quot;pos&quot;:[1360,470],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:166},&quot;flags&quot;:{},&quot;order&quot;:18,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;positive&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;link&quot;:210,&quot;label&quot;:&quot;positive&quot;},{&quot;name&quot;:&quot;negative&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;link&quot;:133,&quot;label&quot;:&quot;negative&quot;},{&quot;name&quot;:&quot;control_net&quot;,&quot;type&quot;:&quot;CONTROL_NET&quot;,&quot;link&quot;:118,&quot;label&quot;:&quot;control_net&quot;},{&quot;name&quot;:&quot;image&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;link&quot;:253,&quot;label&quot;:&quot;image&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;positive&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;links&quot;:[240],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;positive&quot;},{&quot;name&quot;:&quot;negative&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;links&quot;:[239],&quot;shape&quot;:3,&quot;slot_index&quot;:1,&quot;label&quot;:&quot;negative&quot;}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;ControlNetApplyAdvanced&quot;},&quot;widgets_values&quot;:[0.6,0,0.9500000000000001],&quot;color&quot;:&quot;#2a363b&quot;,&quot;bgcolor&quot;:&quot;#3f5159&quot;},{&quot;id&quot;:6,&quot;type&quot;:&quot;CLIPTextEncode&quot;,&quot;pos&quot;:[60.70651182809405,1212.4017172501627],&quot;size&quot;:{&quot;0&quot;:403.20635986328127,&quot;1&quot;:79.98438262939453},&quot;flags&quot;:{},&quot;order&quot;:14,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;clip&quot;,&quot;type&quot;:&quot;CLIP&quot;,&quot;link&quot;:271,&quot;label&quot;:&quot;clip&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;CONDITIONING&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;links&quot;:[133],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;CONDITIONING&quot;}],&quot;title&quot;:&quot;negative prompt负向提示词&quot;,&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;CLIPTextEncode&quot;},&quot;widgets_values&quot;:[&quot;bad quality, (worst quality:1.2), low quality, text, watermark, normal quality, skin spot, bands, bow tie, waist tie, scarf, bad hand, bad anatomy&quot;],&quot;color&quot;:&quot;#322&quot;,&quot;bgcolor&quot;:&quot;#533&quot;},{&quot;id&quot;:115,&quot;type&quot;:&quot;CLIPSetLastLayer&quot;,&quot;pos&quot;:[725,1148],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:58},&quot;flags&quot;:{},&quot;order&quot;:10,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;clip&quot;,&quot;type&quot;:&quot;CLIP&quot;,&quot;link&quot;:276,&quot;label&quot;:&quot;clip&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;CLIP&quot;,&quot;type&quot;:&quot;CLIP&quot;,&quot;links&quot;:[271,272],&quot;shape&quot;:3,&quot;label&quot;:&quot;CLIP&quot;,&quot;slot_index&quot;:0}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;CLIPSetLastLayer&quot;},&quot;widgets_values&quot;:[-2]},{&quot;id&quot;:12,&quot;type&quot;:&quot;SaveImage&quot;,&quot;pos&quot;:[2151,95],&quot;size&quot;:{&quot;0&quot;:385.31402587890627,&quot;1&quot;:678.3333129882813},&quot;flags&quot;:{},&quot;order&quot;:23,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;images&quot;,&quot;type&quot;:&quot;IMAGE&quot;,&quot;link&quot;:130,&quot;label&quot;:&quot;images&quot;}],&quot;properties&quot;:{&quot;ttNbgOverride&quot;:{&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;,&quot;groupcolor&quot;:&quot;#88A&quot;}},&quot;widgets_values&quot;:[&quot;Images\\image&quot;],&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;},{&quot;id&quot;:101,&quot;type&quot;:&quot;ttN text&quot;,&quot;pos&quot;:[62.758970910125238,957.4017172501585],&quot;size&quot;:{&quot;0&quot;:400,&quot;1&quot;:200},&quot;flags&quot;:{},&quot;order&quot;:6,&quot;mode&quot;:0,&quot;outputs&quot;:[{&quot;name&quot;:&quot;text&quot;,&quot;type&quot;:&quot;STRING&quot;,&quot;links&quot;:[215],&quot;shape&quot;:3,&quot;label&quot;:&quot;text&quot;}],&quot;title&quot;:&quot;prompt提示词&quot;,&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;ttN text&quot;,&quot;ttNbgOverride&quot;:{&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;,&quot;groupcolor&quot;:&quot;#88A&quot;},&quot;ttNnodeVersion&quot;:&quot;1.0.0&quot;},&quot;widgets_values&quot;:[&quot;High detail, &quot;],&quot;color&quot;:&quot;#223&quot;,&quot;bgcolor&quot;:&quot;#335&quot;},{&quot;id&quot;:96,&quot;type&quot;:&quot;BatchPromptSchedule&quot;,&quot;pos&quot;:[8.758970910125342,1343.4017172501629],&quot;size&quot;:{&quot;0&quot;:493.5680236816406,&quot;1&quot;:408.9293518066406},&quot;flags&quot;:{},&quot;order&quot;:15,&quot;mode&quot;:0,&quot;inputs&quot;:[{&quot;name&quot;:&quot;clip&quot;,&quot;type&quot;:&quot;CLIP&quot;,&quot;link&quot;:272,&quot;label&quot;:&quot;clip&quot;},{&quot;name&quot;:&quot;pre_text&quot;,&quot;type&quot;:&quot;STRING&quot;,&quot;link&quot;:215,&quot;widget&quot;:{&quot;name&quot;:&quot;pre_text&quot;},&quot;slot_index&quot;:1,&quot;label&quot;:&quot;pre_text&quot;}],&quot;outputs&quot;:[{&quot;name&quot;:&quot;CONDITIONING&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;links&quot;:[210],&quot;shape&quot;:3,&quot;slot_index&quot;:0,&quot;label&quot;:&quot;CONDITIONING&quot;},{&quot;name&quot;:&quot;NEG&quot;,&quot;type&quot;:&quot;CONDITIONING&quot;,&quot;links&quot;:null,&quot;shape&quot;:3,&quot;label&quot;:&quot;NEG&quot;}],&quot;title&quot;:&quot;Batch Prompt Schedule节点变换控制器&quot;,&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;BatchPromptSchedule&quot;},&quot;widgets_values&quot;:[&quot;\&quot;0\&quot; :\&quot;spring, pink flowers, white flowers, cherry flowers, flying flowers\&quot;,\n\&quot;15\&quot; :\&quot;spring, pink flowers, white flowers, cherry flowers, flying flowers\&quot;,\n\&quot;16\&quot; :\&quot;autumn, yellow leaves, brown leaves, flying leaves\&quot;,\n\&quot;29\&quot; :\&quot;autumn, yellow leaves, brown leaves, flying leaves\&quot;\n\n\n&quot;,240,true,&quot;&quot;,&quot;1girl, beautiful,  sunny, trail, park, forest, red dress, &quot;,0,0,0,0,0],&quot;color&quot;:&quot;#232&quot;,&quot;bgcolor&quot;:&quot;#353&quot;},{&quot;id&quot;:116,&quot;type&quot;:&quot;CheckpointLoaderSimple&quot;,&quot;pos&quot;:[724,987],&quot;size&quot;:{&quot;0&quot;:315,&quot;1&quot;:98},&quot;flags&quot;:{},&quot;order&quot;:7,&quot;mode&quot;:0,&quot;outputs&quot;:[{&quot;name&quot;:&quot;MODEL&quot;,&quot;type&quot;:&quot;MODEL&quot;,&quot;links&quot;:[274],&quot;shape&quot;:3,&quot;label&quot;:&quot;MODEL&quot;,&quot;slot_index&quot;:0},{&quot;name&quot;:&quot;CLIP&quot;,&quot;type&quot;:&quot;CLIP&quot;,&quot;links&quot;:[276,277],&quot;shape&quot;:3,&quot;label&quot;:&quot;CLIP&quot;,&quot;slot_index&quot;:1},{&quot;name&quot;:&quot;VAE&quot;,&quot;type&quot;:&quot;VAE&quot;,&quot;links&quot;:[275],&quot;shape&quot;:3,&quot;label&quot;:&quot;VAE&quot;,&quot;slot_index&quot;:2}],&quot;properties&quot;:{&quot;Node name for S&amp;R&quot;:&quot;CheckpointLoaderSimple&quot;},&quot;widgets_values&quot;:[&quot;aniverse_v15Pruned.safetensors&quot;]}],&quot;links&quot;:[[118,70,0,72,2,&quot;CONTROL_NET&quot;],[130,10,0,12,0,&quot;IMAGE&quot;],[133,6,0,72,1,&quot;CONDITIONING&quot;],[206,94,0,93,1,&quot;CONTEXT_OPTIONS&quot;],[210,96,0,72,0,&quot;CONDITIONING&quot;],[215,101,0,96,1,&quot;STRING&quot;],[220,102,0,92,0,&quot;IMAGE&quot;],[221,10,0,103,0,&quot;IMAGE&quot;],[225,104,0,105,0,&quot;MODEL&quot;],[237,5,0,108,4,&quot;INT&quot;],[239,72,1,108,2,&quot;CONDITIONING&quot;],[240,72,0,108,1,&quot;CONDITIONING&quot;],[241,93,0,108,0,&quot;MODEL&quot;],[245,108,0,10,0,&quot;LATENT&quot;],[253,102,0,72,3,&quot;IMAGE&quot;],[257,98,0,112,1,&quot;INT&quot;],[259,100,1,112,2,&quot;INT&quot;],[260,97,0,112,0,&quot;INT&quot;],[261,112,0,108,3,&quot;LATENT&quot;],[262,100,0,113,0,&quot;IMAGE&quot;],[263,97,0,113,1,&quot;INT&quot;],[265,113,0,50,0,&quot;IMAGE&quot;],[266,113,0,102,0,&quot;IMAGE&quot;],[267,98,0,113,2,&quot;INT&quot;],[268,105,0,114,0,&quot;MODEL&quot;],[269,114,0,93,0,&quot;MODEL&quot;],[271,115,0,6,0,&quot;CLIP&quot;],[272,115,0,96,0,&quot;CLIP&quot;],[274,116,0,104,0,&quot;MODEL&quot;],[275,116,2,10,1,&quot;VAE&quot;],[276,116,1,115,0,&quot;CLIP&quot;],[277,116,1,104,1,&quot;CLIP&quot;]],&quot;groups&quot;:[{&quot;title&quot;:&quot;Inputs输入区&quot;,&quot;bounding&quot;:[-84,-8,674,843],&quot;color&quot;:&quot;#8A8&quot;,&quot;font_size&quot;:24,&quot;locked&quot;:false},{&quot;title&quot;:&quot;Outputs输出区&quot;,&quot;bounding&quot;:[2068,-3,535,1798],&quot;color&quot;:&quot;#8AA&quot;,&quot;font_size&quot;:24,&quot;locked&quot;:false},{&quot;title&quot;:&quot;work area 工作区&quot;,&quot;bounding&quot;:[687,-6,1284,1798],&quot;color&quot;:&quot;#b58b2a&quot;,&quot;font_size&quot;:24,&quot;locked&quot;:false},{&quot;title&quot;:&quot;Prompt提示区&quot;,&quot;bounding&quot;:[-85,874,678,921],&quot;color&quot;:&quot;#a1309b&quot;,&quot;font_size&quot;:24,&quot;locked&quot;:false}],&quot;config&quot;:{},&quot;extra&quot;:{},&quot;version&quot;:0.4}</pre><p>After successful loading, use the “Manager” on the right to install the missing nodes corresponding to the workflow and restart multiple times to ensure that the workflow can be loaded properly.</p><p>For users who are not using Tensorbee, please download the relevant models in the workflow on your own:</p><p><a href="https://civitai.com/models/107842/aniverse">aniverse model</a></p><p><a href="https://huggingface.co/guoyww/animatediff/blob/main/v3_sd15_mm.ckpt">animatediff 1.5 model</a></p><p>lcm_sd1.5_lora: <a href="https://huggingface.co/latent-consistency/lcm-lora-sdv1-5/blob/main/pytorch_lora_weights.safetensors">latent-consistency/lcm-lora-sdv1–5</a> put into the loras directory</p><p>Once you have placed the mentioned models and “lora” into the corresponding directory in comfyUI, you can start the generation of the first video.</p><p>ComfyUI should be capable of autonomously downloading other controlnet-related models.</p><p>On a machine equipped with a 3070ti, the generation should be completed in about 3 minutes. The resulting video is presented below:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fbuu-EPZ8IDo%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dbuu-EPZ8IDo&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fbuu-EPZ8IDo%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="640" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/1b976cb14ef1f2abb6488bc52f329ff8/href">https://medium.com/media/1b976cb14ef1f2abb6488bc52f329ff8/href</a></iframe><h3>Progressing to generate additional videos</h3><p>Thanks to the incorporation of the latest Latent Consistency Models (LCM) technology from Tsinghua University in this workflow, the sampling process requires only six steps to generate the corresponding video. Similarly, for a video lasting around two seconds, utilizing 20 steps of sampling on a 3070ti machine, the original 8-minute process is now reduced to 3 minutes with the use of LCM lora. If you need to repeatedly generate the video, the required time is even shorter. Therefore, using this workflow, you can experiment with creating longer videos.</p><p>Given that the video loader currently sets a maximum frame count of 1200, generating a video with a frame rate of 12 frames per second allows for a maximum video length of 100 seconds. The blogger conducted tests using a video lasting 84 seconds.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2Fm-cjc9JVyHk%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3Dm-cjc9JVyHk&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2Fm-cjc9JVyHk%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/eb91e87316be271008e7ce752a45e081/href">https://medium.com/media/eb91e87316be271008e7ce752a45e081/href</a></iframe><p>The ultimately generated video is presented below:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FKzyI1_PQ5mU%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DKzyI1_PQ5mU&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FKzyI1_PQ5mU%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/2e53ca9781ccc189dcee9829be6288be/href">https://medium.com/media/2e53ca9781ccc189dcee9829be6288be/href</a></iframe><p>On a 3070ti machine, the process took 1 hour, which is remarkably fast compared to other approaches to video generation.</p><h3>Introduction to Important Considerations</h3><p><strong>Save in Various Formats</strong></p><p>Within the format settings section of the video combine component, you have the flexibility to adjust the output format of the video. The default is currently set to MP4, but feel free to switch it to GIF, WebP, or any other format that suits your sharing preferences.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/499/1*Nq4QCGcuZOkpQ0ASSCcr3g.png" /></figure><p><strong>Setting Video Width and Height</strong></p><p>In the input area at the top left corner, adjust the width and height of the video based on the aspect ratio of the original video. It is recommended to keep one of the values as 512 and modify the other value to ensure that the generated video maintains the same proportion as the original.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hkymKM7bvT1ANfzCZ1IPhA.png" /></figure><p><strong>Setting the Seed Value</strong></p><p>The current workflow has a fixed seed value. If you wish to experiment for better results multiple times, consider setting the seed value to random.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/394/1*-dUGNbKXZZmhZppDZiPefg.png" /></figure><p><strong>Positive and Negative Prompts</strong></p><p>In the prompt word section, you have the option to adjust both positive and negative prompt words. It’s recommended to append a comma at the end of the prompt words to ensure better integration with the subsequent node batch prompt schedule.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/727/1*0VwSmOSUX08qyqmRifb7Bg.png" /></figure><p><strong>Batch Prompt Schedule</strong></p><p>Within this input box, you have the option to set the starting frame number along with its corresponding scene. For instance, the following settings indicate the utilization of distinct scenes in frames 0 and 16.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/733/1*jiGa40PVl8FN0t0K42pkMA.png" /></figure><p>Please be aware that the input box above requires the use of English quotation marks, colons, and commas. Additionally, it is essential to ensure that <strong>the last line does not end with a comma, whereas other lines must conclude with a comma.</strong></p><p><strong>Controlling Controlnet Weight</strong></p><p>The current workflow assigns a weight of 0.5 to controlnet. It is advisable to set this value between 0.3 and 0.6 for the best video generation results. Avoid setting it too high, as it might impact the quality of the generated video.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/498/1*g6r2EhrXhbDm4B4YKkRudg.png" /></figure><h3>Summary</h3><p>The primary workflow involves extracting skeletal joint maps from the original video to guide the corresponding actions generated by AI in the video. The generation of other content in the video is controlled by prompt words. Therefore, if you bypass the controlNet node mentioned above, this workflow naturally becomes a text-driven video generation workflow.</p><h3>Reference</h3><p><a href="https://stable-diffusion-art.com/animatediff-prompt-travel-video2video/">How does AnimateDiff Prompt Travel work?</a></p><p><a href="https://github.com/luosiallen/latent-consistency-model">latent-consistency-model</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f6cec78f640b" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>