This application's function is to enable the dance move "split." You can use it purely for entertainment or incorporate it into your film and television projects.
I trained this model using a 4090 24G graphics card for about 6 hours, and this is at step 2400. The training utilized a total of 8 vertical resolution materials, so using vertical resolution yields better results. In my testing, it supports full-body photos and clothing like shorts and T-shirts well.
The video displayed on the page was generated using this LORA on a quantized Q8 model with 8 steps. If you use a better base model and higher steps, I believe you can produce higher-quality videos.
You can upload your images, adjust resolution and frame rate to create your own videos!

By registering through my link, you can receive 1000 points, and logging in daily gives you an additional 100 points, allowing you to generate many images and videos for free!

https://www.runninghub.cn/ai detail/1960617834315472898/?inviteCode=rh v1158