Regardless of the oc, make use of concepts included, such as uniform, highheels, jellyfish style haircut.
==============
Norma (Policewoman) | MOSS Image Character LoRA
English
Norma is an adult fictional policewoman: aqua-teal layered jellyfish hair with a rounded short upper bob and longer fine lower strands, green eyes, navy police cap, light-blue short-sleeved uniform shirt, navy tie and pencil skirt, sheer black pantyhose and black pointed stiletto pumps with red soles.
Trigger: nrmnorma
Recommended starting strength: 0.55, especially alongside Meloris at 0.55.
Keep short hairstyle, green-eye and uniform anchors in the prompt. Describe Norma's hands and any held props separately from the other character's restraints.
Prompt anchor:
nrmnorma, an adult woman, aqua-teal layered jellyfish hair, rounded short upper bob with longer fine lower strands reaching the shoulders, clearly green eyes, navy police peaked cap, light-blue short-sleeved uniform shirt, dark navy tie, navy pencil skirt, sheer black pantyhose, black pointed stiletto pumps with red outsoles only underneath.
Known limitations: exact jellyfish layering, green eyes and shoulder length can vary. Spare handcuffs may be duplicated, assigned to the wrong hip or connected by an unwanted long waist chain. In multi-character scenes the adapter can affect other subjects' hair, eyes or uniform details. Exaggerated facial expressions are not reliably reproduced in the current test setup.
中文
Norma 是一名成年虚构女警角色：水青色水母头，上层圆短波波头、下层细长发束至肩；绿眼、海军蓝警帽、浅蓝短袖制服衬衫、深蓝领带与包臀裙、黑色连裤袜、黑色尖头细高跟鞋，红色仅在鞋底。
触发词：nrmnorma。建议从 0.55 开始，尤其是与 Meloris 同框、两支角色 LoRA 各 0.55 的组合。请保留简短发型、绿眼和制服锚点，分别描述她的手部动作和另一人的拘束。
已知问题：水母头分层、发长和眼色会有变化；腰挂备用手铐可能重复、左右不准或多画跨腰长链。多人图中发色、眼色、徽章可能串扰；夸张表情仍不稳定。
训练信息 / Training
MOSS Image Base INT8; KohakuMossTrainer; 48 selected synthetic Anima-generated images per character; natural-language captions; rank 32 / alpha 32; 1200 steps; learning rate 5e-5; AdamW; bf16 adapter training over a frozen INT8 base; native aspect ratio at approximately 1024×1024 pixel area; maximum one image per training batch; no horizontal flip. The text encoder and base weights were frozen. DMD and BunnySlop were not merged into these character adapters.
每个角色使用 48 张筛选过的 Anima 合成图，以自然语言重新标注。训练底模为 MOSS Image Base INT8，rank 32 / alpha 32，1200 步，学习率 5e-5，AdamW；保留原生宽高比、约 1024×1024 等效面积，单批最多一图，不做水平翻转。冻结底模与文本编码器，仅训练角色适配器；未将 DMD 或 BunnySlop 合并到权重中。
Tested inference / 已验证的出图配置
MOSS Image SFT INT8 + qwen3.5_9b_int8_convrot + flux2-vae. Character LoRA: 0.55 each in the two-character setup; BunnySlop_MOSS: 0.5625; MOSS_SFT_8step_DMD_latest: 0.7. 8 steps, CFG 1, Euler / Simple, denoise 1. SkipLayerGuidanceDiTSimple: single_layers=22, double_layers empty, start=0, end=1. Typical duo resolution: 640×1920; street crowd tests: 1024×1024. No artist tags, regional masks, reference-image input or ControlNet in these tests.
当前主要验证于 SFT INT8，而不是同一 Base 训练底模。双角色同框时推荐各 0.55；这不是对所有单人场景的强度最优结论。示例图若搭配 BunnySlop、DMD 或另一支角色 LoRA，会单独注明，不应把整张图的画风都归因于本角色 LoRA。
Credits / 致谢
MOSS Image: https://civitai.com/models/3005406/moss-image
KohakuMossTrainer: https://github.com/KohakuBlueleaf/KohakuMossTrainer
BunnySlop (external style adapter used in the styled examples, not bundled): https://civitai.com/models/2613391/bunnyslop-ai-style-moss-or-anima-or-illustrious?modelVersionId=3412969
File: norma_moss_v1_base.safetensors
Size: 198237768 bytes
SHA256: d3818a701a4473126cde126f28d4b44404da09caf5405e2ba629fb4124b7b421