Instructions to use Viggle/Meridian with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use Viggle/Meridian with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Viggle/Meridian", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
Geometry-only camera rendering on Windows with an RTX 4060 Ti 16GB?
Hi @yycc and the Viggle team,
I'm a 3D/VFX artist interested in using Meridian to change the camera viewpoint of an existing video.
I only need geometry reconstruction and rendering from a new camera path. I’m happy to keep holes or grey regions where source data is missing, so I don’t need the MiniMax-H3 video generation/completion stage.
My setup:
Windows 11
NVIDIA RTX 4060 Ti, 16GB VRAM
64GB system RAM
I noticed the --preview-only option and the Meridian Geometry ComfyUI node. Could you clarify:
Can this geometry-only workflow run on native Windows, or is Linux/WSL2 required?
Can I skip downloading MiniMax-H3, its VAE, and both Meridian LoRAs entirely?
Is 16GB VRAM enough for geometry reconstruction and camera rendering? What clip length and settings would you recommend starting with?
Can the Studio camera-path editor and geometry preview work without loading the video generation model?
Is the Meridian Geometry node available on official Comfy Cloud, or does it currently require a local installation?
If there is a minimal geometry-only installation guide or example workflow, I’d really appreciate a link.
Thanks for sharing Meridian!