This model is a six stage video pipeline for WanVideo 2.2 (GGUF), with post-processing (RIFE VHI, Upscale, ReActor Face Swap). This is the pipeline from the DreamBridge 17 workflow as a standalone workflow.
Takes any image and a set of prompts
Description
First version
FAQ
Comments (6)
Takes about 2 minutes per stage on a 5070 Ti (so 12 mins in total for 28 seconds video @16FPS - before RIFE)
v3.7 takes about 25 minutes on the same config (extra Flux and ReActor passes, plus Lora/Model loads)
"[rgthree-comfy][Reroute] You are using rgthree-comfy reroutes with a ComfyUI Primitive node. Unfortunately, ComfyUI has removed support for this. While rgthree-comfy has a best-effort support fallback for now, it may no longer work as expected and is strongly recommended you either replace the Reroute node using ComfyUI's reroute node, or refrain from using the Primitive node (you can always use the rgthree-comfy "Power Primitive" for non-combo primitives)."
Ah, thank you - I had no idea, I'll swap out the primitives for the rgthree versions in the next version of the workflow - thank you for letting me know!
:/
WanVideoSampler
cannot access local variable 'callback_latent' where it is not associated with a value
ooo.... I haven't seen that from it. Are you using it with a lora? If you let me know how to reproduce it, I'll have a go at figuring out what's up and a fix