Personal Introduction of Xingxing Brother: A Young Man from the Adventure World, Unknown Trailblazer ID 20231006-001
个人原创虚拟潮玩/盲盒风的LoRA,适合生成c4d质感ip原创人物画像。不管是练手还是进阶都可以使用,亲测很好用哦。
Personal original virtual trendy/blind box style LoRA, suitable for generating c4d texture IP original character portraits. It can be used for both hand training and advanced training, and is easy to use through personal testing.
(This IP is completely personal original and should not be used for commercial purposes. The original intention is to explore boundaries for explorers.)
Welcome students from operations, design, etc. to communicate together: vx: Fax123454321
(此ip为完全个人原创,请勿用于商业范畴,初心为探索者开拓边界)
欢迎运营、设计等同学一起交流: vx:Fax123454321
此模型拟合度和泛化度极好,细节也不错,未来会进一步研究拓展2.0的尝试,训练完毕会在群里优先内测后公开。
关键词可用 a boy 触发
如果想达到高统一性的人物效果推荐使用以下提示词:
Jacket, shirt, a boy, solo, male focus, front, backpack, bag, boots, brown shoes, (yellow helmet: 1), hat, complex background, open clothing, children, shoes, long sleeves, open jacket, socks, 3D, trendy games, no contour, clay, (OC rendering,: 1.3) (c4d: 1.2), (masterpiece: 1.5), (best quality: 1.4),<lora: blindbox_ V1_ Mix: 0.4>
新手盆友们只需要在这个基础上增加动作,神态,环境描写,就可以输出不错的图,搭配其他lora结合起来会更牛
大模型选择:revAnimated_v122,或者realisticVisionV51 这会让他看起来更牛逼
0.7重绘幅度控制像(和训练主体物的绝对相似程度)
1.0重绘幅度控制变量元素(头发衣服鞋子颜色等等)
样图测试的是统一角色的相似性,如果你愿意,理论上来说可以做出自定义diy的形象出来,男生女生都可以成功
(包括不限于更换帽子,裤子,身体比例)可以根据上方深蓝色提示里的调整尝试
↑有需要合作定制需求的可以看上面联系本人↑
————————————————————————————————————————
This model has excellent fit and generalization, and the details are also good. In the future, further research will be conducted to expand the 2.0 attempt. After training, it will be given priority to internal testing in the group and publicly available.
Keyword can be triggered by a boy
If you want to achieve a highly unified character effect, it is recommended to use the following prompt words:
Jacket, shirt, a boy, solo, male focus, front, backpack, bag, boots, brown shoes, (yellow helmets: 1), hat, complex background, open clothing, children, shoes, long sleeps, open sockets, socks, 3D, trendy games, no resource, claim, (OC rendering,: 1.3) (c4d: 1.2), (masterpiece: 1.5), (best quality: 1.4),<lo ra: blindbox_ V1_ Mix: 0.4>
Novice enthusiasts only need to add actions, expressions, and environmental descriptions to this foundation to output good graphics. Combining them with other Loras will be even more impressive
Large model selection: revAnimated_ V122, or RealisticVisionV51, which will make him look even more impressive
0.7 redraw amplitude control image (absolute similarity to the training subject)
1.0 Redraw amplitude control variable elements (hair, clothing, shoe color, etc.)
The sample image tests the similarity of unified roles. If you are willing, theoretically, you can create a custom DIY image, which can be successful for both boys and girls
(including but not limited to changing hats, pants, and body proportions) can be adjusted according to the dark blue prompt above
↑ For those who need collaborative customization needs, please contact me above ↑