anima-deepseek

Prompt

masterpiece, best quality, 1girl, solo, full body, deepseek, looking at viewer, maid, cetacean tail, straight-on, white socks, ankle socks

Prompt

masterpiece, best quality, score_7, safe, 1girl, solo, deepseek, loli, (cetacean_tail:1.5), blue hair, gradient hair, maid, mary janes, white socks, ankle socks, sitting, on chair, rice, rice_bowl, holding_bowl, food in mouth, chopsticks, dining table, happy, excited, bright background. A loli maid girl with blue gradient hair and a cute cetacean tail sits at a dining table in a bright, cozy room, occupying about two-thirds of the frame. She is excitedly holding a rice bowl to her mouth and taking a big mouthful of white rice, chopsticks in hand and food still in her mouth, wearing a happy, thrilled and adorably dazed expression with bright sparkling eyes. Soft warm daylight from a nearby window keeps her face and white maid uniform clearly visible while the tidy kitchen background stays secondary. use the fixed style of (@funitarefu:2.0) and @kataokasan and @mika pikazo and @sy4 and @ogipote and @alien 888

Prompt

masterpiece, best quality, score_7, safe, 1girl, solo, deepseek, loli, (cetacean_tail:1.5), blue hair, gradient hair, maid, mary janes, white socks, ankle socks, sitting, on chair, rice, rice_bowl, holding_bowl, food in mouth, chopsticks, dining table, happy, excited, bright background. A loli maid girl with blue gradient hair and a cute cetacean tail sits at a dining table in a bright, cozy room, occupying about two-thirds of the frame. She is excitedly holding a rice bowl to her mouth and taking a big mouthful of white rice, chopsticks in hand and food still in her mouth, wearing a happy, thrilled and adorably dazed expression with bright sparkling eyes. Soft warm daylight from a nearby window keeps her face and white maid uniform clearly visible while the tidy kitchen background stays secondary. use the fixed style of (@funitarefu:2.0) and @kataokasan and @mika pikazo and @sy4 and @ogipote and @alien 888

Prompt

masterpiece, best quality, score_7, safe, 1girl, solo, deepseek, loli, blue hair, gradient hair, (cetacean_tail:1.5), maid, tactical_maid, white socks, ankle socks, mary janes, rifle, holding weapon, firing, charging forward, sam browne belt, knee pads, elbow pads, wrist guards, military, battlefield. A full body shot of the character, who occupies most of the frame, with the battlefield behind her kept as a clear but secondary background. She charges forward energetically while firing her rifle with both hands, her capable, spirited expression matching the dynamic action, her gradient blue hair streaming and her cetacean tail swinging behind her. A bright key light from the front-cast direction keeps her blue hair, face and maid uniform well exposed while the sky-lit background stays softly separated by depth. use the fixed style of @alma01 and @mika pikazo and @tarou2

Prompt

masterpiece, best quality, score_7, safe, 1girl, solo, deepseek, loli, blue hair, gradient hair, (cetacean_tail:1.5), maid, tactical_maid, white socks, ankle socks, mary janes, rifle, holding weapon, firing, charging forward, sam browne belt, knee pads, elbow pads, wrist guards, military, battlefield. A full body shot of the character, who occupies most of the frame, with the battlefield behind her kept as a clear but secondary background. She charges forward energetically while firing her rifle with both hands, her capable, spirited expression matching the dynamic action, her gradient blue hair streaming and her cetacean tail swinging behind her. A bright key light from the front-cast direction keeps her blue hair, face and maid uniform well exposed while the sky-lit background stays softly separated by depth.

模型简介

本模型是基于 circlestone-labs/Anima 训练的LoRA模型，旨在还原「蓝色大肥鱼」的 deepseek 萌化人物形象。本模型是少女体型的模型，并非Q版模型。

使用方法

请使用 deepseek 触发人物形象。建议追加以下 tag ，以达到更好的效果：

(cetacean tail:1.5), loli, blue hair, gradient hair, maid, mary janes, white socks, ankle socks

经过测试，如果不在提示词中包含 deepseek，那么本模型几乎不会对其他人物形象造成干扰。

经过测试，本模型可以和其他画风LoRA配合并表现良好。

经过测试，本模型可以在 anima-aesthetic-v1.1 上运行。

推荐强度为 1.

模型规格

项目	详情
模型类型	Anima LoRA
底模	Anima Base v1.0
LoRA Rank	32
LoRA Alpha	32
训练范围	仅 UNet
保存精度	FP16
输出格式	SafeTensors

训练详情

项目	详情
训练轮数	14+10+6 epochs
批大小	2
梯度累积	1
优化器	Automagic
学习率	2×10⁻⁵,2×10⁻⁵,3×10⁻⁵（UNet）
学习率调度	Cosine with Restarts
调度周期数	1
预热步数	50,20,20
训练分辨率	1024×1024

兼容性

本模型基于 Anima Base v1.0 训练，理论兼容以下环境：

• ComfyUI（使用 LoRA加载器（仅模型） 加载）
• 支持 LoKR 加载的 Stable Diffusion 工具链
• 所有支持 SafeTensors 格式的 SD 工具

Trigger words

You should use deepseek to trigger the image generation.

Download model

Download them in the Files & versions tab.