rh-v1-weapon-combat-lora

δΈ­ζ–‡ RunningHub API RunningHub International RunningHub China

Description

Field Content
Model Name rh-v1-weapon-combat-lora
Model Type LoRA
Platforms ComfyUI / RunningHub / Hugging Face
Author RunningHub-@FOURBUNNY
Original https://www.runninghub.ai/model/public/2094236163150626818
Try online https://www.runninghub.ai
https://www.runninghub.cn/runninghub-api-doc-en/

About this model

Be sure to use the workflow I provided to test this LoRA. The tests and parameters for this version are all based on my workflow. Some third-party workflows or plugin workflows may additionally cause issues such as image blurring, dynamic degradation, and sampling differences; these issues are not caused by LoRA itself. Moving forward, I will no longer troubleshoot or respond to issues such as image blurring and quality degradation caused by the use of third-party plugin workflows.

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 

βš”οΈ Weapon Combat V1

 β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”

A LoRA focused on MiniMax H3 high-speed weapon combat, primarily enhancing cinematic martial arts choreography, weapon interaction, large-scale body displacement, weapon flourishes, heavy strike feedback, and execution actions.

Suitable for swords, sabers, spears, polearms, battleaxes, warhammers, shields, fantasy weapons, and energy weapons, etc.

Compared to the base model, main enhancements: 

β€’ Faster, more continuous weapon combat 

β€’ Clearer Attack β†’ Contact β†’ Deflection β†’ Recovery β†’ Next Strike 

β€’ Better weapon-to-weapon interaction 

β€’ Wide-range swings, rotations, and spatial repositioning 

β€’ Direct transition from weapon flourishes to real attacks

 β€’ Stronger heavy strikes and execution actions β€’ More cinematic weapon choreography performance

 β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β” 

πŸ‡ TRIGGER WORD 

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 

 Weapon Combat LoRA Trigger Word: BUNNY 

 You read that right, it really is BUNNY πŸ‡ 

 β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β” 

βš™οΈ RECOMMENDED STRENGTH 

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

 Recommended Strength: It is recommended to start from 0.7–0.9. You can try 1.0 when stronger action performance is needed, but higher weights may occasionally increase the overall character rotation tendency.

 β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β” ⚠️ V1 NOTES ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Weapon stability and combat logic are significantly improved, but weapon deformation, disappearance and contact errors can still occur in extremely complex scenes or crowded fights.

 β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β”β” 🎁 BONUS ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 

The new workflow includes "H3 Weapon Combat Choreography Director – Zero-Memory AI Template" 

You can directly give the TXT file to ChatGPT / Claude / Gemini / Grok, and then tell the AI your character, weapon, scene, duration, and desired action style. The template already includes weapon choreography, camera movement, spatial movement, weapon techniques, executions, transitions, sound effect design, and "good-looking / bad-looking" prompt contrast examples. Even if the AI has no previous chat memory at all, it can still be used directly.

This repository provides rh-v1-weapon-combat-lora weight files. You can load them on RunningHub.

Finetuned from: minimax-h3

Trigger words: BUNNY

Files

File Size Purpose
jiandou.safetensors 148 MiB LoRA weights

License

Published by RunningHub on behalf of the author. Copyright remains with the author. Follow the original project or upstream license.

Training at RunningHub

Train models on RunningHub: https://www.runninghub.ai/page-model

Run Seedance 2.5 with RunningHub API: https://www.runninghub.ai/call-api/api-detail/2133100000000700025

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support