This is Negative Embedding for SD1.5, Not for SDXL. The concept was to improve quality, such as EasyNegative and veryBadImageNegative.
An12 Version has two embedding file.
an12 (34.0KB)
an12_light (16.0KB)
an12_light is a lightweight version of an12 with only 5 tokens. an12 is the full version.
An11 Version has two embedding file.
an11 (42.9KB)
an11_light (7.57KB)
an11_light is a lightweight version of an11 with only 5 tokens. an11 is the full version.
(In An7,An8,An11 this embedding was created by training at https://github.com/tekitou898009890/stable_diffusion_webui_embedding_estimate.)
(An10,An9 was retrained an8 using the kohya-ss version of textualinversion, and two patterns of TI were merged using embedding merge: one that resulted in a more easynegative result and one that was closer to the model picture.)
Please use it in the "\stable-diffusion-webui\embeddings" folder. Please use this embedding with trigger words of each version (ex.an5) in the NEGATIVE prompt form.
*It can be used with other models, but the effectiveness is not certain.
Description
Please use this embedding with 'an5' in the NEGATIVE prompt form.
FAQ
Comments (2)
What NegativeTI is it?
Perhaps there is a prior competitor, so that comparison seems most important.
If they are not covered, they might be able to be used together, but the face and hands would probably fall apart if they are used together.
I'd like to try it tomorrow when I get some time, but I'd like to see other people's advance reports before I do.
Follow this: https://civitai.com/models/56519
and have a look at the report/comparison of user "freezerain" on the right. You will found it very interesting.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.