Stable Cascade in ComfyUI Made Simple

[@JefHarrisnation]

This was a huge help, especially showing where the models go. Running smooth and producing some very nice results.

[@INVICTUSSOLIS]

Great video. For the first time I was able to run it in comfy.

[@HowDoTutorials]

Glad I could help!

[@mugglepower]

Omgod can't express how much I appreciate that you are telling me which ones to get and where to put them. Confused AF which files to download or where to put them lol

[@AriVerzosa]

Excellent video! I've come across simpler workflows but couldn't get them to work. Yours worked perfectly! Thank you for sharing

[@proterotype]

Well that was an easy subscribe. Looking forward to future videos

[@HowDoTutorials]

Much appreciated!

[@uk3dcom]

Excelent tutorial. It will be interesting to see how Cascade models integrate with other nodes and functionality. 👍

[@HymnAndPraise]

works very well, thank you very much, you are natural for teaching

[@HowDoTutorials]

Thanks for the kind words!

[@pn4960]

It’s a very good channel. I’m sure you’ll get many more subscribers soon

[@squirrelhallowino29]

Your audio quality is very professional, that's such a must for starting channels. What is your setup on that? Sounds like the shure

[@HowDoTutorials]

I actually use elevenlabs speech to speech synthesis. It allows me to get good quality with a mid mic while still putting my personality into the speech instead of just doing text to speech which tends to sound a bit lifeless in my opinion.

[@GES1985]

Can you add additional nodes to separately control background, foreground, subject type, subject pose (controlnet?), subject outfit, subject face, etc

[@HowDoTutorials]

I'll likely explore some of those methods in future videos, but in the meantime I highly recommend experimenting with this node pack: github.com/ltdrdata/ComfyUI-Impact-Pack.

[@AkoZoom]

very easy step by step tuto ! thank you ! But my rtx3060 12Go takes about near 2min for the 4 images and the last (which has en H special) is different also (?)

[@HowDoTutorials]

You may want to try using the lite models or adjusting the resolution down to 1024x1024 to improve generation speed. You may also have better luck using the new models specifically for ComfyUI. Here's an updated tutorial: kzbin.info/www/bejne/fYDRfoubn6ZrfMU

[@AkoZoom]

. @HowDoTutorials Oh yep ty ! then the models are no more in UNet folder.. but in regular checkpoints folder.

[@HistoryIsAbsurd]

Thanks for the vid!

[@chanchanchan1137]

thanks bro!!!

[@workcode2665]

thanks :)

[@Luxcium]

I would like to have seen the connections between the nodes it was a great video thank

[@HowDoTutorials]

I appreciate the feedback! Someone else mentioned that as well and I’ll definitely keep it in mind going forward. Here’s the link to download the workflow if you want to explore it a bit: comfyworkflows.com/workflows/15b50c1e-f6f7-447b-b46d-f233c4848cbc

[@Luxcium]

@@HowDoTutorials I don’t need a link (which is nice tho) I hope you get more free time to make more content it is interesting and I personally love the idea of flow charts a lot and I would love someone who has your skills to explain all nodes and the logic behind them and how the connections are made in different ways to get more workflows and getting insights from people like you is really interesting I understand the possibility of time constraints but I think you should be aware that the stuff you make is really useful and valuable

[@HowDoTutorials]

Thank you for the encouragement @Luxcium. I really enjoy making this content and am grateful people are finding it useful. It is indeed a struggle to find time to make videos at the moment, but I’m hoping to get closer to a weekly release schedule soon. I appreciate your support and will keep your suggestions for content in mind.

[@rovi-farmiigranhermanodela8693]

What about all that videos where they use inpainting tools to edit pictures or to aplay "filters" wich AI can do that?

[@HowDoTutorials]

You can do that with ComfyUI too, though in-painting can be done a bit more easily with AUTOMATIC1111. I don’t have a video covering in-painting yet, but this method can give you something like the “filters” you mentioned: Reimagine Any Image in ComfyUI kzbin.info/www/bejne/eYO4g6d_odmZa5o

[@NorthstriderGaming]

If you show your workflow, please move the nodes more apart so people can see your wiring better. The conditioning part is barely visible so applying these connections is difficult up to almost impossible.

[@HowDoTutorials]

Thank you for the feedback. I generally prefer to keep things tight, but I understand where you’re coming from and will keep it mind going forward. You can find a link to the workflow in the description as well if you’d prefer not to rebuild it yourself.

[@Vectorr66]

I see there is an update, is there a video coming for that? Do we still need this workflow that is on here for comfy? Sorry I am new to comfy and just curious. Thanks!

[@HowDoTutorials]

I’m actually editing the video right now! It’ll cover the new method as well as some techniques I’ve discovered. Should be up by this evening. 😁

[@Vectorr66]

@@HowDoTutorials thanks!

[@evanpop12]

The updated workflow doesn't seem to work for now, I get "model.diffusion_model.input_blocks.0.0.weight". However your workflow works flawlessly, thanks for the tutorial!

[@HowDoTutorials]

Glad to hear mine is working for you!

[@jayrony69]

what is your workflow?

[@HowDoTutorials]

Here’s the link: comfyworkflows.com/workflows/15b50c1e-f6f7-447b-b46d-f233c4848cbc

[@GES1985]

How to add inpainting to finetune images better?

[@HowDoTutorials]

I haven’t done much with inpainting in ComfyUI, but I’ll spend some time with it and put together a video if I come up with something useful. I currently do all my inpainting in Automatic1111 webui as I find it much faster and easier to use with the built in sketch pad.

[@WhySoBroke]

Great video!! Is it possible to do img2img?

[@HowDoTutorials]

Stable Cascade does support img2img as well as some unique ControlNets, but I haven’t gotten it working in ComfyUI yet. I’ll likely post a follow-up video highlighting those methods once I get the workflows together.

[@bobbyboe]

regarding the "update" from "comfyanonymous" - I noticed that he is not using the UNETLoader-nodes and suggests to put the models directly into the folder model... didn`t work for me.

[@HowDoTutorials]

Yeah stability released models specifically for ComfyUI and comfyanonymous released new examples and additional nodes after I released this video. In order to get the updated workflow, you’ll need to use the new models, examples and update ComfyUI.

[@bobbyboe]

@@HowDoTutorials well, your folder structure works fine, also with updated comfy. The other videos on YT so far also use that very same special structure... while in the examples from Comfyanonymous a much more simple structure seem to be used (all Cascade models directly in the model-folder (which did`t work for me). Is it possible that the Comfyanonymous-structure is outdated and maybe they yust did not have the time to actualize?

[@billetandsons]

Seed 1337 represent!

[@Ninjutsu2K]

Depending on each graphics card, how do I know each one of those files I need to download? In my case, the RTX 4080 Super.

[@mugglepower]

you need stage a, b c and the textencoder file, the bf16/lite versions are for cards with lesser VRAM. for the 16 gigs of ggdr6x you got on your 4080 you should be fine with the full size models with the shortest names.

[@Ninjutsu2K]

@@mugglepower ty

[@HowDoTutorials]

I would go for the bf16 models. You can likely run the full size models, but in my experience the quality is not perceptibly different with the bf16 models.