Trigger word : senko-san
Weight : 0.8 - 1
Example prompt : "senko-san, an anime girl, 1girl, solo, looking at viewer, blush, smile, short hair, open mouth, simple background, hair ornament, long sleeves, white background, animal ears, hair between eyes, tail, yellow eyes, flower, :d, japanese clothes, fang, hair flower, wide sleeves, orange hair, apron, animal ear fluff, fox ears, v, fox tail, fox girl, ribbon trim, skin fang, ribbon-trimmed sleeves, miko, fox shadow puppet"
Description
FAQ
Comments (3)
The LORA looks good, but you don't have to put example prompts if you .. don't hide the generation data to begin with.
You're the second LORA maker i've found now who hides their generation data.
With the data inside the images we can take one of the example images, send the PNG info to the generator without a need to copy paste or guess.
Contrary to what you believe, I'm not hiding the generation data, or at least not voluntarily. In a way to not tire myself out completely, I had to get creative with ComfyUI and built a custom workflow and nodes that prep all the things I need in order to publish all these Loras, and a side effect of this is the metadata from these images isn't supported by the site. Making all these loras and generating example images takes a lot of time and I'm already working a full-time job that is completely unrelated to AI, if I don't do things like this to lighten my load, I just would have stopped posting them altogether weeks ago.
Sorry if this answer isn't completely clear, English isn't my first language.
@Silly_Specialist_Ai It was clear.
Based on your response, I suppose that makes sense, and is fair enough. So I apologize for the assumption. However I do have to ask then:
Before transferring, I'm sure the other 'frameworks' still use prompting/LORA/Checkpoints etc, would it not be possible, prior to uploading to copy the positive and negative prompts, toss them in the 'notepad writing tool' so that we can see what checkpoint you used, any supporting LORA's (if any), as well as the CFG scale and other specific settings used to get our images to look like yours?
I see 1 person has already made a very similar image below in the user uploaded ones.
The reason for my original frustration was I only download LORA's that I consider at or above a certain quality, and the closer they are to my ideal mental image of a character, so when I see one that looks perfect, or nearly so, I want my own generations to look just as good. And this comes from an appreciation/admiration of the quality of the LORA uploaded.
Or, in fewer words: The more I like the way a LORA looks, the more I want it to look just like the example images the uploader provides because I find what they made to be that good.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.