A Neco-Arc LORA. Thank me later👌.
You probably want to start your prompt off with this:
1girl, full body, :3, cat ears, chibi, cat girl, cat tail, solo, slit pupils
More or less of the time, it will generate the character in the default Neco-Arc pose, but if you want it to strictly commit to that pose, I've tagged the necessary images in the dataset with "necopose". So use that.
I went through too much cursed Neco art just to make a good dataset, so appreciate this or my suffering will have been for nothing.
70-something images, trained on NAI and tested across multiple models.
Trigger token may or may not be necessary, though I used it just in case. And if this shit is NSFW capable, I never wanna fucking know.
Images in preview may be from past versions! Do not expect full replication!
Images you make with this model and what you do with them are your responsibility!
Description
FAQ
Comments (14)
Finally. The best lora is born. You are truly a hero.
Great work, awesome result!
Now this is gold, thank you for this, now neco arc invasion shall begin in the Ai generated image world
What model did you use for the samples? i searched for Anime_corneosShinra26Mix_v1 and didnt find anything
tysm
neco-arc,cute
Oh god...
Doesnt work at all,
{
"model": "stable diffusion",
"model_weights": "CounterfeitV30_25",
"model_hash": "8f6a079942c4c07c46351a9b8547dce5441e1f8fc45b8049733db6aa8dcb2333",
"app_id": "invoke-ai/InvokeAI",
"app_version": "2.3.4.post1",
"image": {
"prompt": "absurdres, girl, dress, highres, masterpiece, sketch, silly cat, skirt, blush, flat colors, neco-arc, necopose, looking at viewer,, withLora(Neco-Arc,0.75) [worst quality, extra limbs, extra legs, low quality ,white background <easynegative>, desaturated, badhandv3 veryBadImageNegative_v1.3 NG_DeepNegative_V1_T75]",
"steps": 30,
"cfg_scale": 5,
"threshold": 0,
"perlin": 0,
"height": 512,
"width": 384,
"seed": 937758402,
"hires_fix": false,
"seamless": false,
"type": "txt2img",
"postprocessing": null,
"sampler": "k_dpm_2",
"variations": []
}
}
You should include the ":3", "chibi", "slit pupils" in your trigger words here, so CivitAI helper users can one click add them all to prompt with the lora
good point
world doesn't deserve it
Burunyee
A few months of learning how to use stable diffusion led to the discovery of this Lora... I am now at peak form. There is no mountain higher.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.