Pomological Watercolor Redmond LoRA for SDXL

TL;DR

• LoRA for SDXL that replicates the USDA Pomological Watercolor Collection style
• Trigger phrase: Pomological Watercolor
• Trained on SDXL 1.0 with Redmond.AI GPU sponsorship

What This LoRA Does

The Pomological Watercolor Redmond LoRA adapts Stable Diffusion XL 1.0 to produce images in the style of the historic USDA Pomological Watercolor Collection — a archive of over 7,500 watercolor paintings of fruit varieties created between 1886 and 1942. The LoRA extends this aesthetic beyond fruit to anatomical studies of animals, insects, and human anatomy.

All sample generations use the trigger phrase "Pomological Watercolor" combined with anatomical subject prompts, yielding consistent watercolor texture, fine line work, muted earth-tone palettes, and scientific labeling conventions.

Training Details

• Base model: stabilityai/stable-diffusion-xl-base-1.0 (SDXL 1.0)
• Training compute: Sponsored by Redmond.AI
• Trigger word: Pomological Watercolor
• Format: Safetensors (PomologicalWatercolorRedmond.safetensors)
• License: Bespoke LoRA trained license — commercial use permitted with attribution; model weights cannot be redistributed under a different license

How to Use

Diffusers (Python)

from diffusers import AutoPipelineForText2Image
import torch

pipeline = AutoPipelineForText2Image.from_pretrained(
    'stabilityai/stable-diffusion-xl-base-1.0',
    torch_dtype=torch.float16
).to('cuda')
pipeline.load_lora_weights(
    'artificialguybr/pomological-watercolor-redmond-lora-for-sd-xl',
    weight_name='PomologicalWatercolorRedmond.safetensors'
)
image = pipeline(
    'illustrative drawing of a ANATOMY OF A GRAPE, half, pomological watercolor'
).images[0]

Automatic1111 / ComfyUI

Place PomologicalWatercolorRedmond.safetensors in your models/Lora folder. Add <lora:PomologicalWatercolorRedmond:0.8> to your prompt (adjust weight 0.7-1.0). Use the trigger phrase Pomological Watercolor in your prompt.

Recommended Settings

• LoRA weight: 0.8–1.0
• Sampler: DPM++ 2M Karras or Euler a
• Steps: 20–30
• CFG: 5–7
• Resolution: 1024×1024 (SDXL native)

Example Prompts

The model card demonstrates this prompt pattern:

illustrative drawing of a ANATOMY OF A [subject], anatomy, pomological watercolor

Subjects that work well:

• Butterfly (shows wing venation and body segments)
• Human hand (shows bone structure)
• Cat head/skull
• Owl
• Grape clusters
• Apple (cross-section)

The LoRA generalizes to other anatomical subjects while preserving the watercolor illustration style.

Common Errors and Fixes

Error | Cause | Fix | Source

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

Style not appearing | Missing trigger phrase | Include Pomological Watercolor in prompt | HF model card Over-saturated colors | LoRA weight too high | Reduce weight to 0.7–0.8 | Community testing Blurry details | Too few steps | Use 25+ steps with DPM++ 2M Karras | General SDXL guidance

FAQ

Q: Can I use this commercially? A: Yes, the bespoke license permits commercial use with attribution. You cannot redistribute the model weights under a different license.

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

Q: What's the difference from other watercolor LoRAs? A: This specifically targets the USDA Pomological Watercolor Collection aesthetic — scientific illustration with anatomical labeling, not general artistic watercolor.

Sources

• Hugging Face model card: https://huggingface.co/artificialguybr/pomological-watercolor-redmond-lora-for-sd-xl
• Civitai model page: https://civitai.com/models/253270
• USDA Pomological Watercolor Collection reference: https://www.usda.gov/