ArtificialGuyBR

Home / Blog / Training LoRAs

LineAniRedmond LinearManga SDXL LoRA: Manga Lineart Generation

Generate manga-style lineart with this SDXL LoRA. Trigger word LineAniAF, supports detailed/minimalist/colorful/B&W styles. 24k+ downloads.

2 sources cited Training LoRAs

On this page

  1. What This Model Does
  2. Training Details
  3. How to Use
  4. Recommended Settings
  5. Example Outputs
  6. Common Errors and Fixes
  7. FAQ
  8. Sources

LineAniRedmond LinearManga SDXL LoRA

TL;DR

What This Model Does

LineAniRedmond LinearMangaSDXL is a Low-Rank Adaptation (LoRA) for Stable Diffusion XL 1.0 that generates manga-style lineart illustrations. Created by ArtificialGuyBR with GPU compute sponsored by Redmond.AI, it activates with the trigger word LineAniAF and responds to style modifiers in the prompt.

Manga lineart character portrait showing clean black-and-white lines

The model produces clean, high-contrast linework typical of manga art — from detailed character portraits to minimalist sketches and full-color anime-style panels.

Training Details

Property | Value

|---|---

Base model | stabilityai/stable-diffusion-xl-base-1.0 (SDXL 1.0) Trigger word | LineAniAF Trained concepts | Lineart, LineAniAF Style tags | detailed, minimalist, colorful, black and white License (HF) | CreativeML OpenRAIL-M License (Civitai) | Apache 2.0 Civitai downloads | 24,318 Civitai likes | 2,301 The author notes this was their first anime LoRA, trained after being inspired by a popular SD 1.5 manga LoRA.

How to Use

ComfyUI / Automatic1111 / diffusers

  1. Load the LoRA (file: LineAniRedmond-LinearMangaSDXL.safetensors)
  2. Set LoRA strength: 0.7–1.0 (start at 0.8)
  3. Include LineAniAF in your prompt
  4. Add style tags as needed:

Example Prompts

LineAniAF, detailed, manga lineart, beautiful anime girl, intricate hair, sharp focus
LineAniAF, minimalist, black and white, simple character sketch, clean lines
LineAniAF, colorful, dynamic action scene, manga panel, speed lines, dramatic pose

Colorful manga action panel with dynamic pose

The author notes the LoRA "is not perfect and sometimes needs more than one gen to create good images."

Example Outputs

Minimalist manga lineart with clean simple strokes

The three samples above demonstrate the range: a detailed B&W character portrait, a colorful action panel, and a minimalist clean sketch — all from the same LoRA with different style tags.

Common Errors and Fixes

Error | Cause | Fix | Source

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

Weak style activation | LoRA weight too low | Increase to 0.8–1.0 | Author notes Inconsistent results | Single generation | Generate 2–4 variants, pick best | Author notes Style tags not working | Missing trigger word | Always include LineAniAF first | Model card Color bleeding in B&W | CFG too high | Lower CFG to 5–6 | Community practice

FAQ

Q: What base model does this require? A: SDXL 1.0 (stabilityai/stable-diffusion-xl-base-1.0). It may work on SDXL fine-tunes but results will vary.

Q: Can I use this commercially? A: HuggingFace lists CreativeML OpenRAIL-M; Civitai lists Apache 2.0. Check both licenses for your use case.

Q: Does it work with FLUX or SD 1.5? A: No, this is an SDXL LoRA. Civitai lists other base models but the HF card specifies SDXL 1.0 only.

Q: Where can I test it free? A: The author provides a demo space: https://huggingface.co/spaces/artificialguybr/artificialguybr-demo-lora

Sources