Front end to WAN-KR workflow to caption an image via Florence and use WAN2.2 as I2V to create it anew and then feed it to WAN 2.2 for I2V.
You additionally need the low noise checkpoint for WAN 2.2 T2V although it should also work with WAN2.1 T2V
How to use: drop your old dusty SD1.5/SDXL image, select a target resolution for WAN T2V - as high as your machine permits and generate. In case you want to check the result first before I2V block the video gen with the black group bypasser.
If T2I looks fine, unblock GFXCARD_GOES_BRRR and let it do it's thing.
You can add your own prompt directives to the Florence caption, e.g. describe a camera movement, or an action happening.
TO DO: Add possibility to modify/enhance the Florence caption via LLM
Description
Initial release
FAQ
Comments (2)
too complex
well, it's a front-end of 2 simple workkflows (caption and WAN2.2 T2I) to an already complex I2V workflow. If you attach it to a run of the mill single clip generation it will simplify things considerably. Or you can take what you need from it and build your own streamlined wf.
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.