Wan 2.2 Text to Image GGUF Workflow (LightX, Torch Compile, Sage Attention, Nag): v2.0: Wan 2.2 Text to Image GGUF Workflow 2.0.zip

Got rid of the Get-Set Nodes and replaced them with way too many reroute nodes lol.

But for real, ditched the High Noise model after some testing as it is simply unnecessary imo. Changed to Bong Sampler as it produces superior images.

Check "About this version" for what changed or look at the screenshot. Any feedback is more than welcome! I'll maybe try to do a simplified version soon