⚠️使用警告与商业授权声明⚠️
此LoRA模型仅供社区内部进行技术交流与学习研究使用，严禁任何形式的商业用途。
禁止转售或搬运： 未经许可，严禁将本模型文件进行任何形式的转售、分享或上传至其他平台。
禁止商业用途： 禁止将本模型生成的任何图像、视频、动画、游戏资产等，出售、出租、广告宣传、出版物等商业活动。
法律责任： 一经发现上述侵权行为，将保留追究其法律责任的权利。
商业合作： 如有任何商业使用需求，请务必通过联系我进行授权购买。
感谢您的理解与配合。（适用于wan2.1和wan2.2）
(V2.2)
An early version of high-noise training, used in conjunction with the low-noise version. Since high-noise model training is still in the exploratory phase, instability may occur.
(V2)wan2.2 版本发布，wan2.2 version released
wan2.2基于低噪模型训练，使用时需要高低噪模型同时加载，并且高噪权重为2，低噪权重为1.测试环境基于kijai-wanvideo2_2_I2V_A14B_example_WIP修改。
wan2.2 is trained based on a low-noise model. When used, both the high-noise and low-noise models must be loaded simultaneously, with the high-noise weight set to 2 and the low-noise weight set to 1. The test environment is modified from kijai-wanvideo2_2_I2V_A14B_example_WIP.
Prompt word examples
H-gfbz-2.2，女人向前行走，镜头跟随然后慢慢拉远，女人向前行走的同时服装逐渐变成粉白配色的草莓主题蛋糕裙，蓬松的裙摆由三层薄纱叠加，每层边缘缝制奶油蕾丝花边。上衣是纯白方领短袖，胸口点缀立体草莓刺绣与珍珠纽扣，腰间系着红色蝴蝶结缎带。搭配白色蕾丝长袜与红色玛丽珍鞋，头戴草莓发箍，手持迷你草莓蛋糕手拿包，整体甜美如童话里的点心精灵，变身完成后用双手对镜头比心
(V1)wan2.1
这是一个万相图生视频的变化lora，可以进行人物变装，形象变化，例如变动物，还支持首尾帧变化
测试环境 kijai工作流，wan-i2v-14B-720P，wan-vace-Start and end frames
提示词示例
1，在街道上，一位穿着白色T恤牛仔短裙的女人，在向前行走的同时服装变成淡粉色汉服，gfbz效果。
2，宁静的中国寺庙院落，一个穿着袈裟的和尚站在一颗巨大的树下面，和尚向前行走的同时逐渐变成一只大熊猫，gfbz效果。
This is the wan2.1-i2v Infinite Variations Lora.. Currently, it has been tested and supports character costume changes, transformations (e.g., into animals), and also changes between the first and last frames.
Testing Environment: Kijai workflow, wan-i2v-14B-720P，wan-vace-Start and end frames
Example Prompts:
A woman in a white T-shirt and denim skirt is walking down a street. As she walks forward, her clothes transform into a light pink Hanfu. gfbz effect.
In the serene courtyard of a Chinese temple, a monk in a kasaya robe stands under a giant tree. As the monk walks forward, he gradually transforms into a giant panda. gfbz effect.
祝大家玩的开心，I wish you all a happy time,
Description
destiny
FAQ
Comments (24)
how to do multi changed in single video?
You can try to write two more sets of prompt words. The best way is to capture the last frame and continue to generate
color2023 i want to try but idk how to write the prompt, can you help pls?
WildCentaur Follow the prompt word sample format and write the clothing you want to change
i2v因为是通过提示词描述变化效果，所以可以尽量描述详细一些，例如：在户外绿色植物园里，一位穿着黑白色女仆装的女人，女人在植物园里向前行走的同时服装快速变成红黑配色的哥特式洛丽塔风格连衣长裙。吊带设计，上身黑色部分有蕾丝花边和系带装饰，凸显身材曲线。红色裙摆宽大，边缘装饰着黑色荷叶边，且有黑色波浪形线条点缀。搭配黑色长筒袜，袜身有交叉绑带设计，以及黑色带蝴蝶结的短靴。颈间戴着黑色项圈，手上是黑色蕾丝手套，整体造型华丽又神秘。
Google Translate
Because i2v describes the change effect through prompt words, it can be described as detailed as possible. For example: In an outdoor green botanical garden, a woman in a black and white maid outfit quickly changes into a red and black Gothic Lolita style dress as she walks forward in the botanical garden. The suspender design, the black part of the upper body is decorated with lace and laces, highlighting the body curve. The red skirt is wide, the edge is decorated with black ruffles, and there are black wavy lines. Paired with black stockings, the socks have a cross-strap design, and black short boots with bows. Wearing a black collar around the neck and black lace gloves on the hands, the overall shape is gorgeous and mysterious.
各位大拿，不要搬运到lib 谢谢
搬运不可怕, 可怕的是还加密变成它的独家
👍🏻👍🏻👍🏻
A beautiful LoRA that makes a strength from a previous weakness (as the author points out, you take the last frame from your generated video and make it the first frame of the next video segment, and so on). The lack of background consistency becomes a natural part of the scene transition that matches the costume change!
Your final video can literally be as long as you like - generated from a very easy standard workflow.
Of course! If you feel that the last frame of i2v is uncontrollable, you can use the start and end frames, so that lora can provide a change process. The current acceleration lora will have a certain impact on the dynamics
The trigger word is the gfbz效果。 It was also translated during the translation. If it doesn't work, you can try using Chinese.
nipples on the armor
Is there a recommended flow someone can provided a link or similar for? I see mention of using first and last frames to assist with the transition or 'wan-vace-Start and end frames' but aren't sure of a flow to get and try.
Hope this helps，https://civitai.com/models/1816711?modelVersionId=2055870
Example Workflow，https://civitai.com/models/1816711?modelVersionId=2055870
Ignore note links
Works with Wan 2.2!
Just set strength to 2.
Thanks for sharing your experience, I will try it when I have time.
wan2.2 version is on the way
can't wait!!
The 2.2 model is the Low version, correct? The description says to use it with the High version, but I'm not sure if that's coming later, or if I'm supposed to use v1. Are you still training it?
Jellai Used on wan2.2, the high and low noise models are loaded with V2 version at the same time, the high noise weight is set to 2, and the low noise weight is set to 1. English is not my native language. I don't know if you can understand this translation, or you can understand it by looking at the workflow file. The training based on high noise has also been completed, but I didn't get a good effect, so for the time being, you can use the low noise training version. I will also send a version of high noise for everyone to try