This is an advanced workflow designed for those familiar with ComfyUI who want pro grade results. It is node heavy and has numerous options. I encourage you to take the time to learn it. This model does not adapt well to simple versions.
也有中文说明
Instagram: https://www.instagram.com/synth.studio.models/
Buy me a☕ https://ko-fi.com/lonecatone
This represents many of hours of work. If you enjoy it, please 👍like, 💬 comment , and feel free to ⚡tip 😉
Ver 2.0
Updated LoRA loaders to put them before CFG override
added one for Unconditional model
put the option of only using a single model
Quick math tweak.
Ver 1.0 (Beta)
VRAM friendly. Includes Kajai's Optimizer nodes for low VRAM
Modified the instruct to better do NSFW without the "safety filter" warning.
I mean seriously... a swimsuit??
JSON prompt creator
Load an image 🖼️, or...
use the built in LLM, or...
create it from scratch in the prompt builder
Detailer Daemon
Fixed the creator's quality modifier math.
Hi res upscaler.
Seed VR2 (bake small, blow up big!)
🎨 Post Production Suite
Description
FAQ
Comments (17)
thank you for your work
You're welcome 🙏. Please let me know your settings as well as what works and doesn't.
TypeError: Object of type ModelPatcherDynamic is not JSON serializable
Have you uprated comfyui, or are you running on an older machine? This iscaused by dynamic VRAM or a metadata issue
try turning off the "save with metadata" area.
@lonecatone23 updated comfy, disabled dynamic vram, and tried it with metadata off.
@kungfuc945 Send me your log. I'm running the latest comfy. I can't find the issue
@lonecatone23 it's the easy showanything node that feeds the model name to the image metadata node (in the user settings/getsets group) bypassing it makes the workflow work.
@rumohsturluson916 yes
@kungfuc945 Problem fixed. Ver 1.1.1. I messed up.
Loving your workflows so much. With the current ideogram4 wf beta, it seems that Loras are not working. I can get the same loras to work with my own simple workflow based on the comfyui template. Not sure if it's due to some setting or a bug in the workflow.
bad LoRAs possibly? A LoRA loader is a LoRA loader. Maybe try putting one across the other model as well, but the one I did try functioned fine
Move the lora directly after the model and adding it to the unconditional model helps with characters
@ddzines64112 Just the unconditional, or om both?
I have not tried character LoRAs yet. I made this the dayafter it was released
@lonecatone23 my character lora worked pretty good on the main model only, after i placed it directly after the model. The character improved after adding the same lora to the unconditional model.
Thanks for all the great WFs
@ddzines64112 did you have defferent settings?
The reason why I am asking is that I can use a CR LoRA stack and apply it across both models, or two seperate LoRA loaders for each one. What do you think woud be preferable?
@lonecatone23 if a single loader works, that would be ideal IMO.
Try it now