🌌 Rebecca (Cyberpunk: Edgerunners) — Qwen-Image LoRA
This LoRA generates Rebecca from the anime Cyberpunk: Edgerunners.
It was trained and optimized specifically for the Qwen-Image architecture to reproduce her chaotic personality, expressive poses, and vibrant style.
🧩 Recommended Settings
Trigger Word: R3becc4:
Recommended Strength: 1.0
Samplers
res_2m — High quality
er_sde — Faster generation (recommended)
sa_solver — Faster generation
Schedulers
beta57
bongo_tangent
karras
💡 Prompt Tips
It is recommended to start your prompt with the trigger word, followed by a short description.
Example
R3becc4: pale skin, pastel green hair in twin tails, red and yellow cybernetic eyes, pink tattoos
[your prompt]⚠️ Warning / Training Notes
Due to mixed training data, the model may occasionally produce inconsistent cybernetics or tattoo details for Rebecca.
Some images in the dataset contain different stylistic interpretations of the character, which can cause variations such as:
Slightly inaccurate cybernetic eye designs
Tattoos appearing simplified, altered, or partially missing
Small inconsistencies in cybernetic elements
If higher accuracy is required, adding explicit details about the cybernetics or tattoos in the prompt can help guide the generation.
👤 Contact & Support

Discord: hex_wave
Server:
https://discord.gg/bPbAe4MQSH
(Brazilian AI community server — Toca da IA)
Feedback and character suggestions are always welcome.
Support helps with the creation of new high-quality LoRAs.
Description
MMR -> Mixed Multi Resolution
This experimental version introduces a secondary style dataset, aiming to improve OOB quality.