CivArchive
    Hinata Hyuuga LoRA - hinata
    NSFW
    Preview 43573
    Preview 43579
    Preview 43577
    Preview 43576
    Preview 43574

    Hinata Hyuuga LoRA

    Making models can be expensive. Do you like what I do? Consider supporting me on Patreon 🅿️ or feel free to buy me a coffee

    This is after a request on Discord. It's my first successful character LoRA and I learned a lot after making it.

    For the examples I set the weight to 0.6. However the net is not too strong and will allow for a lot of customization (including changing the eyes). In the example I didn't force them, except fort the last one, as you can see from the prompts. Activates with hinata and hyuuga hinata and you can use empty eyes and similar danbooru keywords for the eyes.

    Let me remind you that this is a LoRA network so you need to use it as such with an additional network plugin or something similar.

    How to use LoRA's in auto1111:

    • Update webui (use git pull like here or redownload it)

    • Copy the file to stable-diffusion-webui/models/lora

    • Select your LoRA like in this video

    • Make sure to change the weight (by default it's :1 which is usually too high)

    Description

    FAQ

    Comments (21)

    3418Jan 18, 2023· 1 reaction
    CivitAI

    Hi, Peeledkot explained me a bit how to create a LORA model from scratch, but his explanation uses google collab. How to create a blank LORA model to train on locally?

    Lykon
    Author
    Jan 18, 2023

    I'm also using a remote colab, my gpu is busy with stable diffusion

    Joe_02Jan 18, 2023

    @lykon Hi. Sorry to ask you this, what colab do you use to train the LoRA?

    Lykon
    Author
    Jan 18, 2023

    you can find it in the Project AI discord. I'm not sure I'm allowed to share it elsewhere.

    3418Jan 19, 2023

    @lykon I just joined in, there are 99 pages of search results mentioning LORA :P It's probably truncated to 99.
    What am I looking for, exactly?
    LORA is revolutionary, compared to 8 and 4 GB models! I really want to work with this standard!

    Joe_02Jan 19, 2023

    @alltheirbasesnowbelongtous Could you link the discord?

    Lykon
    Author
    Jan 19, 2023· 2 reactions

    @alltheirbasesnowbelongtous search kohya_ss Trainer Colab in the lora embeddings folder

    3418Jan 19, 2023· 1 reaction

    @pichugeneral2756 Search "Project AI discord" on brave search engine. It should come first in the results!

    Satyam_SSJ10Jan 19, 2023

    It's throwing errors and errors on my PC, mainly the Accelerate

    3418Jan 19, 2023· 2 reactions

    @victorc25 https://github.com/bmaltais/kohya_ss << This is likely what I was looking for. Time to cross fingers and hope it works well on my end!

    AkioAIJan 19, 2023· 3 reactions
    CivitAI

    What model did you use to create the images?

    Lykon
    Author
    Jan 19, 2023

    Midnight Mixer Alt V2 and Abyss Orange Mix 2

    BizoriJan 22, 2023

    @lykon do u have the link for downloading?

    Lykon
    Author
    Jan 22, 2023

    @Sibiridup not on hand now, but they're super easy to find on huggingface. Expecially Orange Mixes.

    BizoriJan 22, 2023

    @lykon I couldnt find Midnight Mixer Alt V2

    Lykon
    Author
    Jan 22, 2023· 1 reaction
    BizoriJan 22, 2023· 1 reaction

    @lykon thx

    AI_HaibaraJan 19, 2023· 3 reactions
    CivitAI

    can it create the before shipuden hinata too?

    Lykon
    Author
    Jan 19, 2023

    You can try. The dataset is scraped

    ai4uiuxFeb 3, 2023· 1 reaction
    CivitAI

    your embeddings are amazing, can you please do tsunade and ino from naruto as well.

    Lykon
    Author
    Feb 3, 2023

    We'll see. Thanks