vid2vid (Easy Peasy) with AnimateDiff, ControlNet, Upscale, Face Swap/Restore, some Filters版本v2.0 (ID: 613694)

ComfyUI NO OneClick Workflow

• Links /Resources

• How to install custom nodes by hand

• Needed Custom Node: https://github.com/Rvage0815/ComfyUI-Rvaged.git

• The Custom Nodes you need are mentioned in the Workflow.

• Keep your ComfyUI updated or some of the nodes don't work

• use the official repo of ComfyUI

描述:

• reuploaded with an optical fix.

• the archive contains 2 versions (anything everywhere and a real linked version)

• i've changed the ControlNet Groups AIO_1 & AIO_2 to Blending wich can be turned on/off so the groups can be used as before but also blend 2 aio preprocessors (like i use most of the time with lineart/hed, sometimes lineart/anyline).

• replaced Depth Anything v1 with Depth Anything v2

• i've removed the LoRA Stack because the loras can be added directly to the prompt so it was a waste of space. (check the comfyui settings in the control panel if loras are activated in text autocomplete. needs custom-script from pythongossss).

• I've also removed Face Restore CF because GFPGAN isn't working anymore with the current ComfyUI Version and Restore Face from reActor gives the same result with codeformer but can also use the GPEN Models (can be found in Links / Resources)

• added Level Adjustments

训练词语: