Try to work with multiple models at the same time, need to further match the weights, the effect is OK<h2 id="recommended-weight:-0.8~1-e9fag1g8a">Recommended weight: 0.8~1</h2><h2 id="recommended-settings-a7ivp3m6n">Recommended settings</h2><ul><li>Sampler：Euler / Euler A</li><li>Schedule type：Simple</li><li>Steps：30</li><li>Hires. fix：R-ESRGAN 4x+ Anime6B</li><li>Magnification：3~3.5</li></ul>

Angewomon Digimon

Multiple models working together can effectively stabilize model workThe video uses wan2.1 i2v 720 FP16. After debugging different configurations, the video has been effectively improved. I found that the prompt words are still a big deal. Improving the quality of the prompt words can greatly improve the video quality.

Files with SHA256: a82247f4afb31120e546f05a547880768510562efbb1f65d9635bbddafa1d71d