Please make sure to use this lora as a negative prompt!!!
(It seems that using a -1 weight in positive prompts can also take effect,Using 1 weight for positive prompts can have an interesting style expression)
Text slider train based on https://civarchive.com/articles/9695/f-best-negative-prompt-for-noobai-xl article,at the same time, some adaptation to Furry Prompt has been added.
After testing, it has been found to improve composition quality and repair hands on various versions of noob(training base model is noob eps10), while having minimal impact on the image of the same seed.
Suggest using this lora on noob.
Description
FAQ
Comments (6)
Isn’t this type of model the sort that should be an embedding rather than a lora?
This Lora is based on https://github.com/p1atdev/LECO Conduct training,
The mechanism is not the same as embedding.
@lse14 I know it’s not an embedding, otherwise it would have shown in the "type" description about this model. I simply expressed my confusion, as to why an "easy negative" model is a lora rather than an embedding.
@GreenMagenta Because this Lora was indeed trained with a large number of negative tags, I didn't think of any better name at the time
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.