This is my long running workflow cleaned up and streamlined for y'all so you get good results using my models out the gate.
There are plenty of notes in the workflow with explanation. I recommend using my SDXL models: https://civarchive.com/models/605283/donutsdeliverymix
Quick Tip: The prompt in Face related section automatically gets included with the main prompt and does not need to go both places.
Hotfix 9 Jan: location for sam model in instructions was wrong and has been changed to /sams/
Help and Questions
Got any questions you can catch me in my discord server: https://discord.gg/kt2hCJHuKp
Description
Fixed for broken comfy nodes
Added Donut Detailer + Donut Text Encoder
FAQ
Comments (36)
Just installed it. Gonna test. Impressed with the amount of things going on in this workflow. lol.
No matters what I do. I cannot make this node work: clip-l vs clip-g (positive).
I have no idea what the hell I'm looking at or how to solve its errors.
uploaded a version that should work out the gate, of course you still need to get all the detection models etc, but download links are in the workflow aswell.
Good morning! Thank you for sharing the workflow, it looks very complete. Unfortunately, I wasn't able to run it. I downloaded the files as instructed in the workflow notes, but nothing happens. It only runs a text file and nothing else.
runs a text file?
@DonutsDelivery Yes, the workflow does nothing. When you press run, it loads the model and that’s it — it finishes. Then it outputs 12 text files. I downloaded all the files you mentioned in the instructions, and that’s all it does. I’m not sure if I did something wrong. But I followed the step-by-step instructions, installed the missing nodes, and downloaded the files mentioned in the instructions. =z
I love this workflow, but I'm getting stuck at the face detailer. Any troubleshooting tips?
How to prompt for clip stuff?
No idea what you mean by that. You are always using clip to encode your text prompt.
Thanks for this workflow, I understood a lot thanks to it!
Just want to say thanks, your efforts are appreciated. I have used your workflows for some time - albeit, I bolted a few extras to it (and so i really appreciated the inclusion of 'layer utility:purge'. I've using your v6 and so far it is a great improvement. One thing I have yet to understand is how to control the weight of the Img2Img. Am I missing something? Thanks either way.
Why does he keep saying I'm missing SimpleMathSlider+。
How can i do?
update comfy essentials to nightly
@DonutsDelivery Still the same error
PLZ
@samo50998 did u solve the problerm, man? I have same
@DonutsDelivery Hi men, thank u for workflow, it insane! But have same problem - Failed to find the following ComfyRegistry list. The cache may be outdated, or the nodes may have been removed from ComfyRegistry. That about comfyui_essentials
@alexeenkostas10835 update the comfy essentials custom node to nightly (version)
@DonutsDelivery Can you provide a screen of the values that need to be in the Simple Math? At least for me I need to just create a new version (this is an issue with some nodes). Problem is I can't see what your original was set to because it doesn't load properly.
P.S. the owner of comfy essentials says they're no longer updating as of April. So that may also be the issue.
If anyone else has this error, don't use comfy essentials
You have to use a fork of it.
@TienX or change version to nightly. it has always worked and still does.
I confirm that switching to Nightly to fix missing SimpleMathSlider+ works.
absolutely awesome results right out of the gate with you model and this workflow!! thank you!! it wouldve taken me hours to try and replicate these results
by adding a force clip to device node from extramodels after the main model speeds things up drastically (about 10x until the first image generation for me). beside this - awesome job!
thank you for this thorough workflow, which by just looking at it is going to teach me a lot .. just one noob question, can you run this workflow on a 2080ti 11GB GPU?? everyone here seems to have 30xx/40xx/50xx GPUs
is there a update workflow ?
Hello, really excited to use this but im missing the "clip-l vs clip-g (positive)" and negative nodes. I saw your previous advice to tell someone to update the essentials (i think) to nightly version and that solved that issue but i havent been able to work out a solution for these 2, so i just disabled them for now in the nodes manager but any sort of help would be appreciated! thanks for the workflow, super detailed!
feel free to disregard this, had to take logic off of nightly. awesome workflow and awesome models! thanks again!
It's crazy how any newer models work with few nodes, yet SDXL requires thousands of nodes to get the result.
newer models give better results from first pass but new models still benefit just the same from upscaling and detailers.
the Math Slide + has a Problem when Insalling the Essentials altough i got the right Version it still tells me its not loaded
i have the same problem, if you know how to resolve them, tell me pls
You have to use this Node Package: ComfyUI_essentials_mb
Hey man, I'm tryna learn I just saw your workflow
I'm hope it works great I didn't try but man you got need to do insane cleanup on your workflow plz plz and subgraphs where it requires and use reroutes, I will wait for an clean worklow update. tnx
I didnt make it for others to use. i made it for myself. and share it because people ask for it.
@DonutsDelivery ohh didn't know that, but still you should use subgraphs they are good.