Special Thanks: This project was made possible thanks to generous sponsorship and GPU time provided by Redmond Ai. We are grateful for their support in training this LoRA.

Model Description

This LoRA imparts a surreal, artistic, and monumental style inspired by the "Balancing Act" series, originally conceived from Google's Veo model outputs. The aesthetic is characterized by large-scale, impossible sculptures that blend technology, classical art, music, and nature in a mesmerizing, often rotating, fashion.

This version is specifically adapted for use with the Tencent HunyuanVideo model.

Trigger Words

The trigger phrase is Balancing Act.

For best results, describe a grand, surreal sculpture or installation, and the model will render it in this unique aesthetic.

Setup for ComfyUI

To use this LoRA in ComfyUI, place the files in their respective directories:

• LoRA: Download Hunyuan-BalancingAct-Redmond.safetensors and place it in ComfyUI/models/loras/

• Diffusion Model: hunyuan-video-t2v-720p-q4_0.gguf (or other Hunyuan model) goes into ComfyUI/models/diffusion_models/

• Text Encoders & VAE: Ensure your clip and vae models for Hunyuan are in ComfyUI/models/text_encoders/ and ComfyUI/models/vae/ respectively.

Example Workflow

You can load this LoRA using a standard "Load LoRA" node in ComfyUI. Connect it between your base model loader and the KSampler. Make sure to set the strength to your desired level (e.g., 0.8-1.0).