Innovate Futures @ Benji
patreon
SkyReels V1 Image2Video Released With ComfyUI! The Fine-Tune Hunyuan Video!
🕑 Added 2025-02-19 19:32:15 +0000 UTC
Video: https://youtu.be/ZGjJG_kT9Ew
Additional Info. Workflows for Patreon Supporters : https://www.patreon.com/posts/122644183/
Skyreel V1, a cutting-edge Hanyuan-based model for high-quality image-to-video and text-to-video workflows! Learn to set up ComfyUI, optimize FP8/BF16 models, and refine outputs with dual sampling and upscaling. Perfect for creating dynamic car commercials, character animations, and cinematic clips—all locally on your PC!
Who is this content suitable for?
AI developers, video creators, and ComfyUI users seeking advanced tools for stable, artifact-free video generation. Ideal for those familiar with Hanyuan Videos or exploring next-gen AI video workflows.
Why does it matter?
Skyreel V1 bridges the gap between raw AI output and professional-grade video, offering smoother motions, reduced pixelation, and creative flexibility. Whether crafting ads, narratives, or experimental clips, this tutorial empowers you to push AI video boundaries without costly subscriptions.
Resources:
Model for ComfyUI: https://huggingface.co/Kijai/SkyReels-V1-Hunyuan_comfy/tree/main
Full Model Weights For Server GPU: https://github.com/SkyworkAI/SkyReels-V1
Basic Image2Video Workflow In this post below.
Have Fun :)
Comments
Aaron
Getting an error when attempting to run it Using a jpg image that is 512 x 512 (cartoon character) # ComfyUI Error Report ## Error Details - **Node ID:** 9 - **Node Type:** UNETLoader - **Exception Type:** RuntimeError - **Exception Message:** Error(s) in loading state_dict for HunyuanVideo: size mismatch for img_in.proj.weight: copying a param with shape torch.Size([3072, 32, 1, 2, 2]) from checkpoint, the shape in current model is torch.Size([3072, 16, 1, 2, 2]).
