Cosmos video model define start and end frame in ComfyUI AI, infinite video, batch processing

Try ComfyUI without hardware restrictions in the RunDiffusion cloud and save 15% the first month:
https://rundiffusion.com?ref=neuron_ai

In this video I show you how you generate videos between a start and an end frame with the local Cosmos video model in ComfyUI. I will show you how you can do this for a whole image sequence and iterate through a whole batch of images in an external folder.

Download this workflow and the base image sequence with the BASE membership on Patreon:
https://www.patreon.com/posts/cosmos-video-and-124059085?utm_medium=clipboard_copy&utm_source=copyLink&utm_campaign=postshare_creator&utm_content=join_link
Find more onfos here:
https://www.nvidia.com/en-us/ai/cosmos/

The needed custom nodes:
https://github.com/BadCafeCode/masquerade-nodes-comfyui
https://github.com/theUpsider/ComfyUI-Logic
https://github.com/Suzie1/ComfyUI_Comfyroll_CustomNodes
https://github.com/WASasquatch/was-node-suite-comfyui
https://github.com/evanspearman/ComfyMath
https://github.com/Kosinkadink/ComfyUI-VideoHelperSuite

Installation of Cosmos:

Update your ComfyUI form the manager.

Download the oldt5_xxl_fp8_e4m3fn_scaled.safetensors model and put it into your folder ComfyUI/models/text_encoders/
https://huggingface.co/comfyanonymous/cosmos_1.0_text_encoder_and_VAE_ComfyUI/tree/main/text_encoders

Download the cosmos_cv8x8x8_1.0.safetensors model and put it the folder ComfyUI/models/vae/
https://huggingface.co/comfyanonymous/cosmos_1.0_text_encoder_and_VAE_ComfyUI/blob/main/vae/cosmos_cv8x8x8_1.0.safetensors

Download the Cosmos-1_0-Diffusion-7B-Text2World.safetensors model and put it in the folder ComfyUI/models/unet/
https://huggingface.co/mcmonkey/cosmos-1.0/blob/main/Cosmos-1_0-Diffusion-7B-Text2World.safetensors

Download the Cosmos-1_0-Diffusion-7B-Video2World.safetensors model and put it in the folder ComfyUI/models/unet/ https://huggingface.co/mcmonkey/cosmos-1.0/blob/main/Cosmos-1_0-Diffusion-7B-Video2World.safetensors

If you're struggling with hardware limitations on your local machine, try ComfyUI in the cloud with RunDiffusion:
https://rundiffusion.com?ref=neuron_ai (affiliate link)

Accessing ComfyUI workflows on RunDiffusion allows you to work with advanced node-based image and video pipelines without installing or maintaining local hardware. By running ComfyUI in the cloud, you can load workflows, experiment with complex node graphs, and scale your work reliably.

https://learn.rundiffusion.com/access-comfyui-workflows-rundiffusion/

Support me on Patreon:
https://patreon.com/neuron_ai

Connect with me:
https://www.instagram.com/_neuron_ai
https://twitter.com/_neuron_ai

#comfyui #stablecascade #stablediffusion #tutorial #animatediff #automatic1111 #aiart #ai #inpaint #interpolation

Видео Cosmos video model define start and end frame in ComfyUI AI, infinite video, batch processing канала Neuron