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.
  • 105 unique users in the last 30 minutes (including lurkers)
    First
    [1]
    		Last
    File: Anima_00157_.png
    (625 KB, 832x1216) ▶
    640888
    No.22319 [Report] AS
    ▶
    I made a checkpoint building up on the older LoRA i made >>21517 . *also I made a new LoRA for more stable output with the new checkpoint.

    i trained it on about 1.5k images what i think is kinda low wich might cause not always perfect images but this is the first finetune i did so far so pls dont expect too much.

    old + new LoRA, previews, checkpoint, and a workflow to get you started in comfyui:
    https://mega.nz/folder/K5g1BZrR#oKeSHzoc8fUFP4cbI7DQnQ

    when i have a larger dataset i will try to make a better version.
    >>
    No.22326 [Report]
    ▶
    Perfect! will there be more guro dataset added? such as works from chunlieater or ayaswan?
    >>
    No.22335 [Report]
    ▶
    >>22326
    Yea currently it was only trained on about 200 images of random guro content
    >>
    No.22340 [Report]
    ▶
    Can you please upload anywhere else? Mega gives "exceeded daily quota" can't download the checkpoint :(
    >>
    No.22342 [Report]
    ▶
    >>22340
    https://gofile.io/d/foK2VO
    >>
    No.22435 [Report]
    ▶
    File: ComfyUI_00973_.png
    (1.66 MB, 896x1152) ▶
    1740118
    >>
    No.23172 [Report]
    ▶
    I've been playing with this and I'm getting like some good results but its not that properly understood, maybe needs to bake a tad harder to understand some concepts. Would you be able to provide a list of tags you've added, stuff like the spit roast or artists?
    >>
    No.23283 [Report]
    ▶
    >>23172
    i never added any artists but next version will have that. all the taggs it was trained on where in the preview images. but next version is jumping from 1.5k images to 4k and i will let the model train a little longer this time.
    [Top]
    Delete post: []
    First
    [1]
    		Last