This is a node setup workflow to compare different textual inversion embeddings in comfyUI.
Use Cases can be comparing of Character likeness embeddings or testing of different strengths of the same embedding.
For testing I am using Emma Watson, Selena Gomez and Wednesday Addams textual inversions, but any other can be put in their place.
If you find this workflow helpful and you use it regularly, consider buying me a coffee:
Description
FAQ
Comments (2)
What custom nodes are you using? I'm getting an error when loading the workflow:
When loading the graph, the following node types were not found:
AdvancedCLIPTextEncode
Nodes that have failed to load will show as red on the graph.
Google lead me here, I replicated the GitHub project in a folder in custmos_nodes, but it's still not loading: https://github.com/BlenderNeko/ComfyUI_ADV_CLIP_emb
well, it is the correct github repo you're at. If the nodes are not working, just replace them with the normal Text CLIP encode prompt node, it will work as well. I just like the advanced one because I like switching prompt strength from confy/a1111/compel