Aesthetic Quality Modifiers - Complete
Training data is a subset of all my manally rated datasets with the quality/aesthetic modifiers, including only the masterpiece, best quality , or very aesthetic tagged images (7273 total).
Superset and final of models in the Aesthetic Quality Modifiers Collection.
ℹ️ LoRA work best when applied to the base models on which they are trained. Please read the About This Version on the appropriate base models and workflow/training information.
Recommended prompt structure:
Positive prompt (quality tags at the end of prompt):
{{tags}}
masterpiece, best quality, very aestheticGeneration Settings:
Previews are generated in Forge with upscaling and adetailer.
For Noobai V-Pred, a ComfyUI workflow reference with DynamicThresholding, Upscaling, and FaceDetailer can be found here: https://civarchive.com/posts/11457095
Description
Trained on Animagine v4 Zero
FAQ
Comments (8)
Hello, I somehow came up with an idea of combining your LORA with a mech, I hope you don't mind, and here's the result.
Hi, I do not mind of course! But I should note that the LoRA version used in your post is trained on Animagine v4 Zero
For illustrious, I have an (earlier) version also trained for this LoRA: https://civitai.com/models/1003636?modelVersionId=1128623
Thanks. :(
I can't believe how much this lora improves my pictures! Great job c: .
So, whats the difference between: masterpiece, best quality, and complete versions of ur loras?
A stronger effect the larger the dataset, as I did not adjust epochs proportionate to dataset size. "complete" having the strongest effect as a result.
Some difference in how the tags were applied:
- Detailed/beautiful compositions and backgrounds for "masterpiece"
- Good anatomy as a baseline for "best quality"
神级lora，简直无敌