The model was created using my personal cloud account after work hours and published during my lunch break. It is intended solely for learning and communication purposes and is prohibited from commercial use.

This lora is intended for the Hunyuan text-to-video generative video model.

This lora depicts the character Lana from blue archive

A girl wearing a dark sailor-style outfit with a white scarf and a choker, who has long white hair, red eyes, and a light pink halo.

The usage of the model is character features + action words to be described.

For example:

A girl wearing a dark sailor-style outfit with a white scarf and a choker, who has long white hair, red eyes, and a light pink halo. A girl is waving.

Workflow JSON:

https://github.com/wochenlong/Civitai_Workflow/blob/main/hunyuan/hunyuan_video_lora.json

Train:

https://github.com/tdrussell/diffusion-pipe

Trained for 10 epochs on a 4090 using 500 static images and natural language descriptions, taking one night.

Comfyui：