READ EVERYTHING BEFORE USING!!!!

So this is a bit of a shill for everyone I got pieces from.

Let's start with the original base I used, created by patreon.com/aitrepreneur 
If you want an easy installer that does most of what this does, check him out as he has a ComfyUI auto installer for his base workflow.

Now for the ugly. I wanted to much apparently. I not only wanted an all in one workflow, but I also wanted...

Wan 2.2 5B Upscaling like what https://www.patreon.com/c/aifuturetech had.

Sage Attention, Florence, and FP16 accumulation like what DigitalPastel has in his https://civarchive.com/models/1847730/smooth-workflow-img2vid?modelVersionId=2091039

Color matching like what GFrost has in his https://civarchive.com/models/1687498?modelVersionId=2082015

This took me 2 days to break everything down and merge the stuff together. Probably would take some of you a few hours, but I never dealt with trying to make, modify, or merge workflows before. So this was one heck of a learning experience, especially how things are routed with DigitalPastels workflow vs this one.

So why did I do it? Because I was blown off when I asked for these things to be included in the workflow so someone had to do it and I'm sure some of you want this.

INSTALLATION!!!

aitrepreneurs auto installer is nice and all for the basic version he offers for free on his patreon but it does not have Sage attention in it and when I tried installing Sage attention in his ComfyUI I never could get it to work. So good news for you....

His auto installer is awesome and it's free! Yes! Just follow along in his video and create a new ComfyUI that DOESN'T require redownloading your models. It has an auto patcher to have multiple ComfyUI installations use the same models. Watch the whole video, it's awesome.

So this is ALMOST a complete suite. Yes there are a few things that prevent me from saying this is a complete suite. First, there is no looping. Why? Because I'm spoiled okay. I am so used to Framepack and what it can and can not do that most of the looping workflows are unimpressive to me except one....

https://civarchive.com/models/1836348/wan-22-i2v-simple-multiple-promptvideo-loop

See, the others are good and all but you lose a level of control when looping. If I want to just have two people walking in the park talking for a while then use a lip sync voice over ai, the mentioned workflow is great. But the others are typically limited to only 4 prompts and what happens if the second prompt goes to crud? I'd rather feed in each 8 second prompt myself, which is why I also included END FRAME IMAGE GENERATION! So now you can feed your stuff in yourself.

I can't help, I showed you what I did and how I got it running. I don't know your system.

Beyond that, butcher this thing like I have done. Make it bigger. Add more stuff on it. Make a badder one so that this one is forgotten.