This model was generated with about 25 cherry picked images of shinobu, I did not upscale any of the images shared on this post and the Lora does a good job most of the time to create the character. I used the abyssOrangeMixV2NSFW checkpoint to train the model and to generate the images seen above, this model is intended to be used for SFW content.
I would very much appreciate if you could leave any reviews/ comments on this post. I will do my best to reply to them all and use it to train my future models. This also give others a chance to see the legitimacy of post
Some trigger words are: Oshino_Shinobu, Donuts, Summer Hat, sitting down
You really only need the first one, but the lora has memory from the other trigger words as well.
Please visit my Pixiv if you want to support me more: https://www.pixiv.net/en/users/19376687