📝 Description

Hey everyone!

I actually created this LoRA a few months ago and finally decided it was time to share it with the world! It's only the second style I've posted here, so I'm really excited to see how it performs out in the wild.

This is an art style LoRA heavily inspired by the amazing artist Dondeboys. It was trained on the Illustrious base model and works beautifully for getting that distinct, high-quality anime/manga aesthetic.

Please feel free to drop your generations, leave feedback, and share your reviews in the comments. Seeing your images and hearing your thoughts genuinely helps me improve, so don't be shy!

Have fun experimenting! ✨

⚙️ How to use it (My Settings)

I usually run this LoRA at a strength of 1.0, but honestly, feel free to play around with the weight to see what works best for your prompts! Here is the exact setup I used during my test runs:

• Trigger Word: dondeboys

• LoRA Weight: 1.0 (experiment as you like!)

• Best Checkpoints to try: waiHASSHAKUMILK_v10 or waiillustrious_v170

• Sampler: Euler a

• Schedule type: Automatic

• Steps: 32

• CFG Scale: 6

• Resolutions(my test): Works great at 1024x1536 (portrait), 1536x1024 (landscape), or 896x1152.

For crisp details, I highly recommend using Hires.fix:

• Upscaler: 4x-UltraSharp

• Upscale by: 2

• Denoising strength: 0.45

Prompt Setup:

• Positive: masterpiece, best quality, highly detailed, sharp focus, dondeboys, <lora:dondeboys_style_10:1.0>, [add your character or scene here]

• Negative: worst quality, low quality, bad anatomy, bad hands, blurry, monochrome, bad eyes (feel free to tweak the negative prompt to fit your needs!)