Video tutorial here -
1 - Shows how to use impact bridge and node templates
2 - Shows an updated quick technique for cut / paste / blend
3 - Shows a quick clipseg model to mask based on text description
INSTALL
Mile High Styler - https://civarchive.com/models/119246/sdxl-mile-high-prompt-styler-now-with-25-individual-stylers-each-with-1000s-of-styles-and-you-can-add-custom-styles-infinitely
Efficiency Nodes - GitHub - jags111/efficiency-nodes-comfyui: A collection of ComfyUI custom nodes.- Awesome smart way to work with nodes!
Impact Nodes - GitHub - ltdrdata/ComfyUI-Impact-Pack
Mikey Nodes - GitHub - bash-j/mikey_nodes: comfy nodes from mikey
Allor Node - GitHub - Nourepide/ComfyUI-Allor: ComfyUI plugin for image processing and work with alpha chanel.
Description
FAQ
Comments (2)
I just try to your mask using words workflow, but i use this wokflow,
1. In apply ipadapter node 'clip_vision_output' input not present, in your video in apply ipadapter have total 4 input (model, ipadapter,clip_vision_output & mask), but in my apply ipadpater node have total 5 input (ipadpater, clip_vision, image, model, attn_mask) present, not present 'clip_vision_output' input. I have uptodate ip-adapter plus node & model, still not present.
2. in my comfyui models folder, 'clipseg' folder not present and CLIPSeg Model Loader not load any model. now tell me what should i have to do?
Hi there - yes, attention mask just came out (after I created this resource/video). I believe it's optional, so you don't need to put anything there. For the models, you need to make sure the open_clip_pytorch_model file is in the clip_vision folder and the clipseg folder from the archive is in the clipseg folder (both of these are under the comfyui/models folder.
Hope that helps and good luck!