Remade-AI/Zoom-Call

Instala las dependencias y carga el LoRA sobre Wan2.1 14B T2V. El prompt debe incluir la frase de activación [z00m_ca11] y describir la posición y la acción de cada participante.

pip install -U diffusers transformers accelerate

import torch
from diffusers import DiffusionPipeline
from diffusers.utils import export_to_video

# Cambia a "mps" para dispositivos Apple
pipe = DiffusionPipeline.from_pretrained(
    "Wan-AI/Wan2.1-T2V-14B",
    dtype=torch.bfloat16,
    device_map="cuda"
)
pipe.load_lora_weights("Remade-AI/Zoom-Call")

prompt = "The video shows a [z00m_ca11] with four participants. In the top left box, a medieval knight in full armor adjusts his helmet. To his right, a pirate with a parrot on his shoulder drinks from a mug. In the bottom left, a scientist in a lab coat scribbles on a whiteboard. In the bottom right, an alien in a suit waves awkwardly."

output = pipe(prompt=prompt).frames[0]
export_to_video(output, "output.mp4")

Ajustes recomendados para ComfyUI: intensidad del LoRA 1.0, escala de guía integrada 6.0 y Flow Shift 5.0.