Skip to content

Jojocodex

minimax-h3-wushu-action-lora

Jojocodex/minimax-h3-wushu-action-lora

minimax-h3-wushu-action-lora is a text-to-video model from Jojocodex. Use it when you need video from a text prompt. It is set up for minimax-h3. The card lists the license as apache-2.0.

武打动作 LoRA,基于 MiniMax-H3(Comfy-Org/MiniMax-H3)训练,让模型生成人体武打/功夫动作(拳、腿、踢、棍法等),专注人体打斗动作优化。 个人在测试中4步和8步加速LORA偶尔会有画面糊的情况。Wushu-Action-LoRA 0.5的强度搭配0.3强度spatial-physics-lora,打斗的效果会更好。

Downloads · 30 days

3.1K

19% of all-time downloads

All-time downloads

16.6K

Public

Repo size

775 MB

Likes

70

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors465 MB · 100%

At a glance

Task
Text-to-Video
Library
minimax-h3
License
apache-2.0
Access
Public
Created
Aug 16, 2026
Updated
Aug 30, 2026
SHA
2fcd16cf

Base models

Task
Text-to-Video
Library
minimax-h3
License
apache-2.0
Wavespeed
live
Created
Aug 16, 2026
Updated
Aug 30, 2026