ArtificialGuyBR

Home / Blog / Training LoRAs

ColoringBookRedmond V2: SDXL LoRA for Coloring Book Art

Generate consistent coloring book line art with this SDXL LoRA. Trigger words, settings, and usage guide.

2 sources cited Training LoRAs

On this page

  1. What It Does
  2. Base Model & Training
  3. Trigger Words
  4. Recommended Settings
  5. How to Use
  6. Example Outputs
  7. Common Errors and Fixes
  8. FAQ
  9. Sources

ColoringBookRedmond V2: SDXL LoRA for Coloring Book Art

TL;DR

What It Does

ColoringBookRedmond V2 is a Low-Rank Adaptation (LoRA) for Stable Diffusion XL that generates coloring book–style illustrations — clean black-and-white line art suitable for printing and coloring. The model was created by ArtificialGuyBR using GPU compute sponsored by Redmond.AI and trained on a large dataset of coloring book images.

Base Model & Training

Note: The Civitai entry (model 209889) is labeled "2.1V for FreedomRedmond SD 2.1" — that is a separate SD 2.1 version. The Hugging Face repo artificialguybr/ColoringBookRedmond-V2 is the SDXL variant documented here.

Trigger Words

Include these tokens in your prompt to activate the style:

ColoringBookAF, Coloring Book

The HF widget uses exactly this prompt. The Civitai page lists the same two triggers.

Parameter | Recommendation

|---|---

Base model | SDXL 1.0 (stabilityai/stable-diffusion-xl-base-1.0) LoRA weight | 0.7 – 1.0 Sampler | DPM++ 2M Karras / Euler a Steps | 20 – 30 CFG | 5 – 7 Resolution | 1024×1024 (SDXL native)

How to Use

ComfyUI

  1. Load ColoringBookRedmond-V2.safetensors via Load LoRA node
  2. Set strength_model to 0.8
  3. Prompt: ColoringBookAF, Coloring Book, <your subject>, clean line art, white background

Automatic1111 / WebUI

  1. Place the .safetensors in models/Lora
  2. Add <lora:ColoringBookRedmond-V2:0.8> to your prompt
  3. Use the trigger words above

diffusers (Python)

from diffusers import StableDiffusionXLPipeline
import torch

pipe = StableDiffusionXLPipeline.from_pretrained(
    "stabilityai/stable-diffusion-xl-base-1.0",
    torch_dtype=torch.float16,
    variant="fp16"
).to("cuda")

pipe.load_lora_weights("artificialguybr/ColoringBookRedmond-V2")

prompt = "ColoringBookAF, Coloring Book, cute cat, clean lines, white background"
image = pipe(prompt, num_inference_steps=25, guidance_scale=6).images[0]

Example Outputs

The model produces consistent black-and-white line drawings with clean contours, minimal shading, and white backgrounds — ideal for coloring book pages. Subjects include characters, animals, objects, and scenes.

Common Errors and Fixes

Error | Cause | Fix

|---|---|---

Colored output instead of line art | LoRA weight too low or missing triggers | Increase weight to 0.8+, ensure ColoringBookAF, Coloring Book in prompt Blurry / low detail | Too few steps or wrong sampler | Use 25+ steps, DPM++ 2M Karras SDXL base not found | Pipeline expects SDXL base | Use stabilityai/stable-diffusion-xl-base-1.0 exactly

FAQ

Q: Is this the same as the Civitai "2.1V" model? A: No. The Civitai model 209889 is for Stable Diffusion 2.1 (base model SD 2.1). This Hugging Face model is for SDXL.

Q: Can I use this commercially? A: The LoRA inherits the CreativeML OpenRAIL-M license from SDXL. Check that license for commercial terms.

Q: What LoRA weight works best? A: 0.7–1.0. Start at 0.8 and adjust.

Sources