all / g / s / fur / art / ai / 3dcg / f / rp / req / p2p / dis / lit / kaki chat
[Settings] [Home] [Catalog] [Search] [Thread list] [Stats] [Reports] [Watcher] [PMs] [Admin]
[Return] [Bottom]

Posting mode: Reply
BBCode
(for deletion)
  • Allowed file types are: gif, jpg, jpeg, png, webp, webm, mp4
  • Maximum file size allowed is 25000 KB.
  • Images greater than 255 * 255 pixels will be thumbnailed.
  • 148 unique users in the last 30 minutes (including lurkers)
    First
    [1]
    		Last
    File: ComfyUI_04658_.png
    (1.09 MB, 1024x1024) ▶
    1144465
    No.8940 [Report] AS
    ▶
    Sorry, first post, hope I didn't do anything illegal.
    So, my fellow degenerates, there is somewhat new model called Chroma. It's basically flux without any limitations, and with additional training on E621 and danbooru datasets. Currently it's not finished, but it's already makes great pictures with natural language prompt.
    Sadly, my English is a-Ok at best, so I'm not very good with it, but it's still can do guro without any additional tweaks.
    It works with ComfyUI, and if I'm not doing anything wrong right now you can use this picture as a workflow example.
    If you have any prompts you want to try, tell me and I will try in spare time.
    Oh, btw, let's not make too much noise about it, model is not done yet (as I mentioned) and I have no idea what it's creator thinks about guro.
    >>
    No.8944 [Report]
    ▶
    I'm waiting for chrom to finish the training, before trainig LoRA on it.

    it is useless as it is. I have my failed datasets that didn't work on FLuX .
    >>
    No.8985 [Report]
    ▶
    IIRc currently chrom is around epoch 31.

    it is supposed to go to 50, and only yhe last 3 will be trained on High res images, 1024, all preveious ones were 512.


    SO untill than I hope and pray.
    [Top]
    Delete post: []
    First
    [1]
    		Last