Showcase Images were created with comfyui in PNG and include metadata, prompt & workflow. All licensing conditions for the Flux Dev model apply. You can find the details here.
Description
Model inspired by Riccardo Guasco artwork.
Trained on Civitai with a dataset of 77 images
Recommended resources : Fluxmania IV
Recommended settings : dpmpp_2m / sgm_uniform / 25 steps / flux guidance : 3.5
Weighting : 0.8
FAQ
Comments (9)
I wanted to create a LoRA model for this style, but high-resolution image sources were scarce, thnx.
there is no need to train on high resolution images. Flux can handle 512x512. In my training image databases, I often mix resolutions and it does not affect the quality of the LoRA
@Adel_AI what the name of this style ?
@AbdallahAlswa80 Hi, what does your question refer to, I don't know what you're talking about
@Adel_AI the style ,,, is there a style called ai_colab !
@AbdallahAlswa80 I just created a model from images from an Insta account called The AI colab. I have no idea who it is. I couldn't find any information on the web. I have given the account link in the model information
cool
"...and include metadata, prompt & workflow"
Thank you!!! This should be the standard, but often lora creators obfuscate things for reasons I can't understand.
it's true that it should be more frequent, I found that it helped, and encouraged those who are new to comfyui.