My ComfyUI workflows for T2I/I2I
They are optimized for my checkpoints and created of my latest knowledge to enhance the outcome.
"If this workflow leveled up your day, I'd purr-eciate a like! 😻"
Versions & Information👇👇👇👇👇👇👇👇
👉 Please read below and the file descriptions "About this version" for more info's.
⚠️ Do not use the workflows with the "Nodes 2.0 beta" from ComfyUi or it will mess up things.
👇👇👇👇👇👇👇👇
What you get from the comfy workflows:
♨️ Easy controls
✅ As less as possible dependencies
🪧 Detailed documentation
⛓️ Highly automatic logic
✨ Optimized results
🎬 Fully automated resolution logic
🔖 Bookmark-Shortcuts with number keys
Types of workflows
DarkArts C-TI2I
🖼️ T2I and I2I
🧩 Automatic aspect‑ratio calculation and fitting
✨ Multiple Upscalers + Refiners
Upscale with Model (additional detail, high quality)
RTX Super Resolution (+ Deblur + Denoise)
⏩ Multiple sampler passes
💎 Easy Quality prompt option
🎰 Multiple Radomierzers & Wildcards
🔖 Bookmark-Shortcuts - with number keys
🩻 Known issues and advice's
none
🫶 Special thanks @TrashAI for the idea and help with the wildcards!
All older Versions are available inside my GitHub Repo.
YOU are responsible for outputs as always! If you make ToS violating content and I get aware I WILL report this.
Description
Changes
Enhanced Upscaler nodes
Multi-Pass fixes
Anima + Krea2 compatibility
FAQ
Comments (32)
im so used to seeing t2v/i2v i never paid attention to t2i/iti . i take it those are image creating?
Yes, image creating
Clean workflow as always, very appreciated - only thing I notice is that the image doesn't export with meta data - is there a way to toggle that on?
I'm working on that part ^^
Hell yeah I been checking everyday for this loool
Hahaha 🤣 thank you, hope it helps, but it is not that advanced yet
Hello! Your workflow looks amazing! I really enjoy using your Wan workflow, but I'm a bit stuck here: I would love to use this setup with Illustrious-based models. Since they don't require a separate text encoder (CLIP), how can I properly bypass that node and configure the workflow? I'd really appreciate your guidance!
No clue how I should explain that here. You rewire the clip connections and disable the extra clip loader. Or you just download the clip for illu and use that.
@Darksidewalker so this does work with your illu model then? that's good news. I have been looking for 30 minutes for a workflow from you for illu lol
@ambersai Not out of the box, changes would be needed. I normally use SwarmUI for image gen, that's why there is no wf.
@Darksidewalker Well obviously some changes will need to be made of course. Is it better to use the illu clip? Is it one you tweaked, or the standard clip for "not AIO" illu?
@ambersai I did not alter any clip, there is no reason to
@Darksidewalker ok, just the standard one then. I will find it. I didn't know if maybe you had made your own for your illustrious merges.
I would like to ask what the differences are between this workflow and these two workflows:
https://civitai.com/models/2581985/dasiwa-hybrid-workflows-or-i2v-or-flf2v-or-i2v2v-or-audio-or-app
They both seem to be for Wan Video 2.2.
Could you explain the differences between them?
One is just wan22 the others SVI
you should categorize this correctly, not under wan2.2. i assumed i could gen an image from the t2i capability of wan when its a totally different model altogether
It is categorized correctly 🙄
Hmm, the checkpoint in this workflow looks for it in the actual checkpoint folder as opposed to unet like most in ANIMA.
I do hope he changes this to either unet or, more preferably, to use the diffusion models folder, because this isn't a merged model, and keeping it in my checkpoints folder is going to confuse me later if I want to use it with another workflow.
Haha, I like it! I like to place the model in the checkpoint.
You can just replace that node with a node that looks in the correct folder
Thank you for your existence, i wish you had unlimited money
Wildcards don't work; a wildcard node is needed before the concatenate node for them to function.
Yeah they are mixed, got no time to fix that.
Fixed that some time ago, but forgot to post it here XD
Interesting toggling wildcards method. Will add it to my workflow because i was searching for something like that at some point.
I will stay with Illustrious because Anima don't work with "KSampler (inspire)" that has seed variation and didn't find any other KSampler that has this function working with Anima.
If someone has a solution for that i will perhaps switch to Anima one day.
This old function from A1111 should be a standard on KSamplers. I don't understand why they forgot something this important.
Hi! Thanks for sharing this amazing workflow.
I'm running into an issue with the latest version.
Error:
[WARNING] invalid prompt: Node 'ID #290:827' has no class_type. The workflow may be corrupted or a custom node is missing.In the workflow, the DaSiWa_TorchResize node appears as UNKNOWN.
I'm using the latest ComfyUI and all custom nodes are fully updated via ComfyUI Manager, but the problem still persists.
Is there an additional repository, a newer version of ComfyUI-DaSiWa-Nodes, or another dependency required for this workflow?
Thanks!
I had the same issue, just do to your manager and update Dasiwa nodes package and it should work
How to switch off I2I?