The training and publication of this LoRA were conducted entirely by an autonomous Agent equipped with multiple embedded skills and tools. The complete workflow, including data generation, training configuration, model training, result organization, and automated release, was executed without human supervision. Based on the author’s manual inspection of representative samples, the resulting LoRA demonstrates strong character consistency, stable visual identity, and well-defined character anchors. These observations suggest that the proposed agent-driven workflow is both efficient and technically robust.
Further implementation details are available in the accompanying technical article: https://civarchive.com/articles/32228/single-image-character-lora-training-via-synthetic-expansion
Shian / 诗岸 Synthesizer V 公式服 LoRA for WAI v17 / Illustrious-SDXL.
Trigger words <code>shian_synthv, synthesizer v, 1girl, formula_dress</code>
Recommended Usage
<pre><code>- Recommended LoRA strength: start from `0.70`; usable range `0.55-0.70`.
- This package uses the Forge-evaluated recommended epoch `6`.
- Use the shared character trigger together with exactly one outfit trigger.
- Do not mix outfit triggers in one prompt unless intentionally blending outfits.
- Showcase images in this package follow the v5 rule: each outfit must appear exactly once in a near shot and once in a mid-shot with the full body fully visible.</code></pre>
Outfit Prompts
Synthesizer V 公式服 / <code>official_formula</code>
Outfit trigger: <code>formula_dress</code>
Full prompt
<pre><code>shian_synthv, formula_dress, 1girl, solo, blonde_hair, long_hair, twin_tails, braided_hair, hair_rings, bangs, ahoge, pink_eyes, white_dress, black_bodice, orange_trim, off_shoulder_dress, detached_sleeves, frilled_skirt, layered_skirt, diamond_pattern, orange_thighhighs, mismatched_legwear, black_ankle_boots, black_bow, orange_bow, hair_bow, hair_ornament, hairclip, choker, neck_bow</code></pre>
Example with LoRA
<pre><code>&lt;lora:shian_synthv_strict_imagegen_recommended_e000006:0.70&gt;, shian_synthv, formula_dress, 1girl, solo, blonde_hair, long_hair, twin_tails, braided_hair, hair_rings, bangs, ahoge, pink_eyes, white_dress, black_bodice, orange_trim, off_shoulder_dress, detached_sleeves, frilled_skirt, layered_skirt, diamond_pattern, orange_thighhighs, mismatched_legwear, black_ankle_boots, black_bow, orange_bow, hair_bow, hair_ornament, hairclip, choker, neck_bow, upper body, looking at viewer, soft smile</code></pre>
Negative Prompt
<pre><code>low quality, blurry, bad anatomy, bad proportions, deformed hands, extra limbs, extra fingers, disfigured, low resolution, text, watermark, logo, signature, multiple people, extra character, duplicate person, cropped head, cropped face, out of frame, bad composition, distorted body, long neck, bad posture, asymmetrical eyes, malformed eyes, smeared face, low detail face, out of focus face, inconsistent outfit, different outfit, missing accessories, extra accessories, background clutter, object blocking face</code></pre>
Recommended Generation Settings
<pre><code>- Checkpoint: `WAI-Nsfw-Illustrious-17`
- Showcase resolution: `1024x1024` native output; no Hires.fix, APISR, or other resolution enhancement was used.
- Sampler: `Euler a`
- Steps: `28`
- CFG scale: `6.0`
- Seed used for draft showcase/eval images: `2026073001`</code></pre>
License / Usage NSFW generation is allowed. Author credit is required: jifengwang. Generated images may not be sold. Selling this model, merges, or redistributed model files is not allowed. Use on Civitai and other generation services is allowed. Sharing merges and using different permissions on merges are allowed.

Shian / 诗岸 Single-Outfit LoRA

Recommended epoch 6 for WAI v17 / Illustrious-SDXL. Use weight <code>0.70</code>; tested useful range <code>0.55-0.70</code>.Trigger words: <code>shian_synthv</code>, <code>synthesizer v</code>, <code>1girl</code>, <code>formula_dress</code>.

Files with SHA256: 981213f090ba8b096bc9b835d539ff8f288cc2afec6ba90c79e6160222cfe7bb