ArtificialGuyBR

Home / Blog / Training LoRAs

StoryBookRedmond V2: SDXL LoRA for Children's Book Illustrations

Generate consistent children's book style images with StoryBookRedmond V2 LoRA for SDXL. Trigger words, settings, and usage guide.

2 sources cited Training LoRAs

On this page

  1. What This Model Does
  2. Trigger Words and Style Control
  3. How to Use
  4. Recommended Settings
  5. License
  6. Common Errors and Fixes
  7. FAQ
  8. Sources

StoryBookRedmond V2: SDXL LoRA for Children's Book Illustrations

TL;DR

What This Model Does

StoryBookRedmond V2 is a Low-Rank Adaptation (LoRA) trained on top of Stable Diffusion XL 1.0 (stabilityai/stable-diffusion-xl-base-1.0). It specializes in generating illustrations that look like they belong in children's storybooks — soft, whimsical, and narratively composed.

The model was created by ArtificialGuyBR with GPU compute sponsored by Redmond.AI. Training used a large curated dataset of children's book imagery. On Civitai it has ~1,500 downloads and on HuggingFace 428 downloads (as of August 2026).

Trigger Words and Style Control

Include these in your prompt to activate the style: Trigger | Purpose

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

KidsRedmAF, Kids Book | Core activation — required detailed | More intricate line work and shading minimalist | Clean, simple compositions colorful | Vibrant, saturated palette black and white | Monochrome line art style Combine triggers for mixed effects, e.g., KidsRedmAF, Kids Book, detailed, colorful.

How to Use

ComfyUI

  1. Load StoryBookRedmond-V2.safetensors via Load LoRA node
  2. Set strength_model to 0.8 (start) — range 0.7–1.0
  3. Add trigger words to your positive prompt

Automatic1111 / WebUI

  1. Place LoRA in models/Lora/
  2. In txt2img, click the LoRA tab and select it
  3. Add <lora:StoryBookRedmond-V2:0.8> to prompt (or use the UI picker)
  4. Include KidsRedmAF, Kids Book in prompt

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/StoryBookRedmond-V2")
pipe.fuse_lora(lora_scale=0.8)

prompt = "KidsRedmAF, Kids Book, a brave mouse knight, detailed, colorful"
image = pipe(prompt).images[0]

Parameter | Value

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

Base model | SDXL 1.0 LoRA weight | 0.7–1.0 Resolution | 1024×1024 (native SDXL) Sampler | DPM++ 2M Karras / Euler a Steps | 20–30 CFG | 5–7 The author notes: "LORA is not perfect and sometimes needs more than one gen to create good images." Expect to reroll occasionally.

License

CreativeML OpenRAIL-M — same as the base SDXL model. Allows commercial use with attribution and sharing of derivatives under the same license. See the HuggingFace model card for the full license text.

Common Errors and Fixes

Error | Cause | Fix | Source

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

Style not appearing | Missing trigger words | Always include KidsRedmAF, Kids Book | Model card Overcooked / burnt look | LoRA weight too high | Reduce to 0.6–0.7 | Community feedback Inconsistent characters | LoRAs don't guarantee character consistency | Use ControlNet + IP-Adapter for characters | General SDXL knowledge

FAQ

Q: Does this work with SD 1.5? A: No. This LoRA is trained specifically for SDXL 1.0.

Q: Can I use this commercially? A: Yes, under CreativeML OpenRAIL-M. Provide attribution and share derivatives under the same license.

Q: What's the difference between V1 and V2? A: V2 was trained on a larger dataset with more compute (Redmond.AI sponsorship). The author recommends V2.

Q: Why do I need multiple generations? A: LoRAs add style bias but don't guarantee every seed works. Reroll with different seeds.

Sources

Back to all models · SDXL LoRAs · Children's book styles