<h1 id="v1-u3o4f2pzv">V1</h1><p><span style="color:rgb(193, 194, 197)">Workflow for Wan 2.1 text to video that can generate a small resolution video first, then upscales it when you find one you like. The workflow has teacache and sage attention to reduce generation time, but sage does requires </span><a target="_blank" rel="ugc" href="https://github.com/woct0rdho/triton-windows?tab=readme-ov-file"><span style="color:rgb(34, 139, 230)">triton</span></a><span style="color:rgb(193, 194, 197)"> to run.</span></p><p><span style="color:rgb(193, 194, 197)">If you don't want to use sage attention or don't have triton installed, you can just bypass the node.</span></p><h1 id="v2-sfrr7atcn">V2</h1><p>Changed the workflow slightly to make it faster to run while keeping quality.</p>

Wan t2v 2step generation

<p>First version with teacache and sage attention.</p>

Files with SHA256: 901d1a055173d5b8f5d513f28a96f0e9934429a431a2a6880f945a6f220a45e7