CivArchive
    Hellcoder - (Trying to uncensor Stable Diffusion 2.1) - Alpha_Hellcoder_V_0-3
    Preview 1534129
    Preview 1534130
    Preview 1534131
    Preview 1534134
    Preview 1534133
    Preview 1534135
    Preview 1534136
    Preview 1534137

    Ever noticed how hard was to train the 2.1 model and its derivates? It was because it had a heavily censored text encoder, which is the thing that makes pairs with images and roughly understands what's inside the pictures you provide to the model.


    Well, there are better and uncensored text encoders compatible with the 2.0-2.1 models on huggingface, and once you get the model you want to train on in diffusers format, you can competely change the single parts of the model. Now, that's what I have done with this model and I will keep working on it.

    I have trained the model on a large amount of women of any kind, using the waifu diffusion tagger, and by doing this the model captured the single features of each woman and mixed them together to get info for the celebrities that the encoder knows about but the unet is not trained on.

    Currently the model has low capability on NSFW stuff, but I will try to make it work next, until then you can have some acceptable depictions of celebrities, I very well know these photos aren't perfect but it is a huge step forward on the basic model.

    Ah, since it is not overfitting, it should be able to deal with other concepts, try yourself and, since it was smartly merged with Illuminutty you can try with their embeddings...

    Description

    (The smaller file is the Negative Embedding)

    This particular version seem to work its very best at 512*512, I mainly trained it at this resolution for the lowest end video cards (it works on my 4G quite fast). Despite that, it generates decent 768*768 images too!

    FAQ

    Comments (6)

    Redigleb_Doppler2482
    Author
    Jul 13, 2023
    CivitAI

    New version uploaded, the next will feature a fine-tuned negative prompt for anyone who wants to use the model in Diffusers (that does not support embeddings). This is useful if you want to use it on colab without WebUIs (which technically are NOT accepted by Google terms of service)

    GetModels2023Jul 13, 2023ยท 2 reactions
    CivitAI

    See if you can get an attractive female sucking a man's dick. All these portraits that are G-rated make me chuckle.

    settima_aiJul 13, 2023

    Probably you are asking too much for now, but if you look at relative improvements in such few time....

    Redigleb_Doppler2482
    Author
    Jul 13, 2023

    It is already possible, but you would not want to post porn pictures in a thread focusing on celebrities, do you? I already posted a "NSFW dedicated one"... which is kind of the same model, I'll be updating (the model I already uploaded focuses on nudity alone)

    HV3Aug 11, 2023

    stick to p*orn google searches or SD1.5 - SD2 released with specific methods to reduce NSFW outputs so do a bit of digging to find out why most of SD2.x is mostly SFW outputs.

    seedmancOct 26, 2023

    censorship is not all about nsfw, you know

    Checkpoint
    SD 2.1 768

    Details

    Downloads
    719
    Platform
    CivitAI
    Platform Status
    Available
    Created
    7/13/2023
    Updated
    4/30/2026
    Deleted
    -

    Files

    hellcoderTryingToUncensor_alphaHellcoderV03.pt

    hellcoderTryingToUncensor_alphaHellcoderV03.safetensors

    Available On (1 platform)

    Same model published on other platforms. May have additional downloads or version variants.