ComfyUI_ExtraModels
ComfyUI_ExtraModels copied to clipboard
[Bug] Encoding seems to be faulty when batch size of latent > 1
Strange behaviour if batch size >1. It is the same with gemma on cpu or on cuda...