Is it possible to add a feature that identifies/highlights the prompt for each element in the generated image?

[tjw2469]

My idea is that it would be possible to visualize what works and what does not work in your prompt and what has a negative impact. It might also help to use the negative prompt effectively to avoid obvious flaws in the image, while reducing the number of prompts that pile up.

[Mousewrites]

Yup!

https://www.youtube.com/watch?v=XiKyEKJrTLQ

Script is in the extension tab.

[tjw2469]

Thanks a lot.
The installation was easy, but when I test it, it just shows me the error:
"TypeError: FrozenCLIPEmbedderWithCustomWordsBase.tokenize_line() takes 2 positional arguments but 4 were given"
do you know the cause by any chance?

[Mousewrites]

add a comma between every word in your prompt in the script box (i think)

[tjw2469]

I tried, it does not work

[Mousewrites]

i dunno then. Not my tool, just something that does what you wanted. Might be other daam repos out there.

[tjw2469]

I found the problem. The DAAM directed from Stable Diffusion's built-in extension index URL was no longer working, I removed it manually and replaced it with a new version and now it works!