🐙 OctopussMixS Style LoRA ✨
I've been sitting on this one for a while. This is the style LoRA I distilled from my own OctopussMixS generations — over 1,000 images — and used to build OctopussMixA v2. Figured it was time to just release it 😄
🎨 What it does
Pushes toward that anime 2.5D zone I've been chasing across all my models. Illustrated but with depth, painterly but not flat, skin that actually responds to light. The aesthetic I've been refining for a while now, packaged into a LoRA. It also makes your images more colourful.
🔧 Usage
Trigger: ```@octopuss```. Load it, set a weight, generate. Start around 0.7–0.8 and adjust from there.
📌 Honest note
This was originally trained to be baked into a checkpoint rather than used standalone, so consider it a look behind the scenes of how I build my models. It works as a standalone LoRA too though — just manage your expectations and experiment freely 💕
Feedback always welcome. Enjoy 🐙