Since CivitAI is going down THE DRAIN
I won't be updating this workflow here anymore, and I won't post any more models from now on.
I hope to find interest from people to keep updating this. Thanks everyone!
Keep following this workflow development and more on my new Discord server: https://discord.gg/EwKE8KBDqD
Get new releases from GitHub: https://github.com/diodiogod/Comfy-Inpainting-Works
Now there is a "simple/compact" version that I will update together with the normal one.
Read "about this model" on the right panel please!! I recommend reading v6.0, v5.0, v5.1 as well.
This is a proper inpainting with alimama control-net + Flux Fill + Flux Tools (Depth an Canny) support from V7 and above.
I try my best to avoid hiding the workflow flow, while balancing keeping it clean and user-friendly.
This workflow will make sure your image is composited in the end to prevent VAE degradation AND from v5 will make sure your image is a multiple of 8 (so it doesn't get resized by ksampler) at the same time it avoids resizing at all (unless you want to) since resizing will degrade the original image as well.
For inpainting that leaves a face and changes the whole background and character, Alimama works way better. Alimama also works better for LoRas. Alimama is terrible at outpainting and Flux Fill is amazing. You should test both. My comfyUI OOM or crashes when switching from one to the other IF my system memory page file is not large enough. With a pagefile of 50GB I can freely switch from Flux Fill to normal Flux Dev now.
It does use a lot of custom nodes to make life easier and workflow cleaner. Use manager to install the missing ones.
Supports loading of an external mask (need to match the image size) or doing the inpainting within the right-click "Open MaskEditor" on the image itself.
Supports FILL -the normal fill, like Auto1111- (use FastBypasser - black node)
Supports area inpaiting x full image inpaiting (use FastBypasser - black node)
Supports variable denoise (Change denoise on the CONTROL ROOM)
Supports Negative (use FastBypasser - black node)
Proper composite so no VAE degradation
Support for Flux Tools (Depth an Canny) with Alimama
I added LoRa support and Daemon Detailer support (both are optional)
Description
I cleaned up some necessary custom nodes. Mainly the Logic-utils that had a bunch installation problems with some people. No more animatedif or video nodes (some of their nodes were being used for simple tasks, but I replaced them.)
Also, "fixed" the view. I lower all nodes to fit the default view. Some people were not seeing any nodes because of that.
I hope that people will experience fewer issues now with the installation.
FAQ
Comments (11)
I tried removing the "Repeat Masks 🎥🅥🅗🅢" from video helper suit, but the "CR repeater" did not work to make a batch of masks... so this will still need to be a mandatory custom node so the workflow works for batches. Sorry.
I tried this workflow but I have a missing "ReplaceString" any tips to fix this? thanks!
Oh I'm sorry, I thought I got rid completely of LogicUtils, but there is this missing node you mention. I'll change it to another more common string replacement. Sorry! I'll fix it today. It's in the end, where it saves the civitai metadata, if you want to change it yourself for now.
New version now should not have this problem 4.1
This needs a video tutorial lol
Yes it does =(
I'm not much of a video guy, but I'll consider
The Checkpoint Name Selector nodes assume we will use the full Flux checkpoints (which actually most don't) the nodes point to the default checkpoint folder in ComfyUI. Most people use the Unet versions of the Dev and Fill Flux models which are in the Unet folder along with the GGUFs.
You should be able to change it to the ones you desire to use. Those loaders are just the default, but if your model is in another folder, I suggest you make a symbolic link inside the checkpoint folder to the unet folder. That is what I do. OR move the models on the unet folder to the checkpoint folder, as I think now, comfyui fully supports Flux loading from there.
And if nothing like that works for you, You can remove this node and add the Load Diffusion node you normally use and connect to the "Switch any" model selector, but you might have to make other adjustments to get the model name to the image-saver (or do it mannually) in the end (to get it to save the model name in the civitai metadata)
I'll think of another way of getting the name to the image-saver node... I used to use "Widget To String" node to get the name from the checkpoint selector, but some other people had problems wth this node so I removed it. If you have another suggestion to help, please do!
I changed this on the new version. See if it works better for you.
@diogod Thank you very much! I can now select the models. It is a very advanced workflow.