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.
  • 103 unique users in the last 30 minutes (including lurkers)
    First
    [1]
    		Last
    File: FLUX_00384_.jpg
    (31 KB, 1152x896) ▶
    32522
    No.3247 [Report] AS
    ▶
    A new Ai txt2img model for creating pics came out in August 2024.

    Flux, it is NOT censored, and it can be run locally. It performs much better than SD1.5 and SDXL. It's on par with Midjourney and Dall-E 3, but it is local, so no prompt is censored.

    You can try it for free here: https://huggingface.co/spaces/black-forest-labs/FLUX.1-dev

    It can be run on as little as 4GB of VRAM (it is tricky to do but possible), so if you have run SDXL or SD1.5 or pony you can run this one.


    I have made some LoRAs for this model. A hanging model and a model where a girl is sucking the barrel of a gun.

    I'm still learning how ot properly make LoRAs so those are versions 0.5 I will make a better version in the future.


    1. Hanging LoRA, trained mostly on 3d images, so it makes everything look 3D, weight 0.5-1, trigger words hanging, hang, noose


    https://files.catbox.moe/oa69r0.safetensors


    2. Gun in the mouth LoRA, sometimes the gun is not drawn properly.


    https://files.catbox.moe/qyyppb.safetensors



    Both have around a 50% success rate. They combine well with the character LoRAs.

    Have fun, and if you use them, please post results.

    If anyone has a specific fetish, please provide a dataset and i will train it.


    It doesn't do anime well yet. SO I will post only the 3D pic here.
    >>
    No.3249 [Report]
    ▶
    what are the tags for gun in mouth?
    >>
    No.3250 [Report]
    ▶
    >>3249
    Her lips are parted, and a black handgun, held by a male hand, is being inserted into her mouth


    She is sucking the barrel of that gun.

    this phrases. first one is better. modify as you see fit
    >>
    No.3251 [Report]
    ▶
    Hi, I'm very interested in an electric chair LoRA. I've only dabbled in AI and have never trained one, what sort of images would you need for an electric chair dataset? I.e., would you rather I provide (fake) realistic electric chair images or anime ones or a mixture? Do they all need to have the same resolution or aspect ratio? How many images is preferrable? And are there any other specifications? If you let me know, I'm more than happy to scour the internet and my own collections and make any nescessary adjustments, and I'd really appreciate a decent electric chair LoRA! :)
    >>
    No.3258 [Report]
    ▶
    >>3251
    If you can give me a dataset I will glady train a lora for it.

    it costs 2.5 USD but I just want some good guro Loras done.
    ,
    Better to have a mix, if you have realistc even better. with, resolution is not important, preferably aorund 20-30 pics. The more the better, at least it was for previou smodels

    I'm learning but i believe FLUX can manage to train a concept with 30 images
    >>
    No.3264 [Report]
    ▶
    Alright, here's over 100 3D and realistic images of girls strapped into the electric chair. I've named thm all "lctrcchr" as a keyword but haven't captioned them past that, as I have no idea how to caption. If you want me to add captions, I'm willing to spend time and learn and provide the dataset again. If there's anything else you need me to change or do just let me know! I will do anything to make this happen haha :D

    Here's the dataset: https://files.catbox.moe/89s755.zip
    >>
    No.3265 [Report]
    ▶
    >>3264
    The dataset is very good. I will caption it and try to train on it. You should probably get an answer in a couple of days
    >>
    No.3266 [Report]
    ▶
    >>3265

    Great, thanks! I AI-upscaled it which I thought was a good idea but I never realised how much it fucked with it in some images. If that causes weird artefacts or something of the likes, that could be the problem. But anyway, thank you so much! I hope it turns out good :)

    Oh, and a couple of the images need rotated like 90 degrees, I forgot to do that.

    Very excited for this!!
    >>
    No.3274 [Report]
    ▶
    >>3266
    this is how FLUX makes the elctric chair with no trainig.
    https://files.catbox.moe/blwv9f.zip
    >>
    No.3275 [Report]
    ▶
    >>3274

    or this
    https://files.catbox.moe/j9ah0l.png
    >>
    No.3284 [Report]
    ▶
    That last image is looking good! Pretty decent model right out of the box! ^^
    >>
    No.3331 [Report]
    ▶
    I'm new to trying to use AI, you mention that Flux can be run locally, which sounds like a good idea to me. But eh... how might one do that? I went to the page and was able to do the free trial thing, but I couldn't find any option to download a program and I didn't want to just stark clicking buttons until something happened xP
    >>
    No.3333 [Report]
    ▶
    >>3331

    Honestly I just followed a tutorial to instal something called "Comfy UI" which lets you use AI models locally, and then follow a tutorial on installing FLUX (dev or schnell version, it's up to you) into Comfy UI.

    The Comfy UI interface might be tricky to get the hang of if you haven't used AUTOMATIC1111 before, but you'll get the hang of it i'm sure :)

    I'm not OP btw, they might have more or different advice.
    >>
    No.3338 [Report]
    ▶
    >>3333
    >>3331

    What this guy said.

    I use COMfyUI (there are easier alternatives to start, but COmfyUI is notbetter long run, so you might as well learn it)
    .
    So find a tutorial on how to install ComfyUI
    https://www.youtube.com/watch?v=Zko_s2LO9Wo

    and then add FLUX to it

    https://www.youtube.com/watch?v=ImWHS5Ux36E

    https://www.youtube.com/watch?v=DdSe5knj4k8
    >>
    No.3340 [Report]
    ▶
    >>3333
    >>3338
    Awesome, time to get learning it seems! Thanks for the help :)
    >>
    No.3357 [Report]
    ▶
    File: FLUX_01855_.png
    (1.31 MB, 1024x1024) ▶
    1378243
    >>3251
    >>3264
    I made the electri chair model.


    testing it now to find the best paramters to use it.


    WIll post with instruction after moer test.
    >>
    No.3360 [Report]
    ▶
    >>3357

    First impression looks great, man! I'm super excited to use it. :D
    Should I post if I make anything good with it?
    >>
    No.3369 [Report]
    ▶
    File: FLUX_02080_.png
    (977 KB, 1024x1024) ▶
    1001041
    >>3251
    >>3357
    I finished my electri chcair LoRA

    it is here
    https://files.catbox.moe/9nlaoh.rar

    with images (imahes include COmfyUI workflows) and instructions.

    have fun. Post results.
    >>
    No.3372 [Report]
    ▶
    >>3369
    Password, man. ;(
    >>
    No.3374 [Report]
    ▶
    >>3372
    damn forgot

    password i 1234
    >>
    No.9042 [Report]
    ▶
    >>3247
    Is that Black Widow?
    >>
    No.9046 [Report]
    ▶
    Can i use this in forge? The only .safetensors model i found was the 23 gig dev model.
    >>
    No.9047 [Report]
    ▶
    >>9042
    yes. glad you brought this thread back to life.


    >>9046
    ... yes?

    You need a flux mode, lighter version exist,

    And the LORA


    Also at the end of the day FLux is not king, i still use SDXL most of the time.
    >>
    No.9056 [Report]
    ▶
    >>3369
    Is there a way you could just upload your stuff here?

    I tried downloading that rar thing on my phone. It said cannot open rar files, and offered me to search the playstore. I downloaded a rar zip app, but it started acting weird like it would mess up my phone.
    >>
    No.9065 [Report]
    ▶
    >>9056
    I cannot directly uplaod it on this website. and catbox is havign monetary issues.

    what format and and what site do you prefer?


    P.S. how the fuck do you run image generative AI on your phone?
    [Top]
    Delete post: []
    First
    [1]
    		Last