WAN 2.1 Video Generator Better Than Hunyuan | workflow скачать в хорошем качестве

WAN 2.1 Video Generator Better Than Hunyuan | workflow

A very powerful and promising model for video generation has been released, which is available locally. After long tests and experiments with the schemes, I can say that it is the best today of what is currently available to us locally. The video is generated consistently, the arms and legs do not fall off, everything looks very decent. Today I will show you two very flexible schemes with a screwed-on Triton, which speeds up generation by 20-30% without losses. Imag 2 video works from 12 GB, and Text 2 video generally from 8 GB and with colossal speed. I have never felt such positivity from video generators on the local network. This model is very promising. MENU 00:00 - WAN 2.1 overview 02:48 - Online service for video generation wan 2.1 text 2 video 05:35 - WAN 2 generation speed 08:03 - ComfyUI installation 08:29 - Installation of models for wan 2.1 11:01 - WAN 2.1 text to video (txt2video) workflow 13:11 - How to write prompt for video generation 15:29 - Triton group in WAN 2.1 scheme 20:51 - VAE Decode - if there is not enough vram 22:09 - WAN 2.1 image to video (img2video) workflow 🔥 My structured course on "AUTOMATIC 1111" with my support on the site https://stabledif.ru 🔥 My structured course on "ComfyUI" with my support on the site https://stabledif.ru/comfyui 🔑 All links and files are available on Boosty: https://boosty.to/stabledif ✔️Telegram channel: https://t.me/stabledif_lesson ✔️ Telegram CHAT: https://t.me/stable_dif ✔️VK Prompts: https://vk.com/stabledif