Method and Thoughts - Miniature People LoRA for Flux - fluxgym

Поделиться
HTML-код
  • Опубликовано: 10 дек 2024

Комментарии • 35

  • @KLEEBZTECH
    @KLEEBZTECH  2 месяца назад +4

    Miniature LoRA: civitai.com/models/766577

  • @SouthbayCreations
    @SouthbayCreations 2 месяца назад +1

    Fantastic video, a lot of good solid information. The lora came out great!! Thanks for the shout out!!

  • @b.s.design
    @b.s.design 2 месяца назад +2

    I downloaded this beautiful model the day it was released, and today I'm already watching your tutorial )) Beautiful work! Thank you so much!

    • @KLEEBZTECH
      @KLEEBZTECH  2 месяца назад +1

      Thanks for watching!

  •  2 месяца назад +4

    I am not expert neither, but I was thinking: maybe if you try to come up with nonrreal trigger word, because when you write "miniature person" it triggers all the neural paths responsible for plastic miniatures. Maybe if you switch the word for "m1n1ppl" or something like that, and avoid using miniature in your prompt, it might pop out better results, closer to your lora training data. But as I said, I am just guessing :D
    Edit: Oh. I am now just on 10:00 and I see you did that :D I should comment only after watching full video next time :D
    PS: You just earned a subscriber :3

  • @Martin-bx1et
    @Martin-bx1et 2 месяца назад +2

    I really appreciate your talk-through on this.

  • @Nalestech
    @Nalestech 2 месяца назад +2

    Super helpful! Thank you!

  • @thewebstylist
    @thewebstylist 2 месяца назад

    Wowww BIG UPS brah!! 🎉👏🏻💯

  • @jmagick7281
    @jmagick7281 2 месяца назад +1

    Works good. I have a realistic renders after 5 tries. Using Mini-People and Faetastic-Details at a strength of 1.0. Would love to see a V2 of this cool Lora. Thanks! Liked and Subscribed.

    • @KLEEBZTECH
      @KLEEBZTECH  2 месяца назад

      Thanks! I do hope to improve it soon. Working on more source material to allow for more options. I keep trying different detail LoRAs but in my testing I was not fully satisfied with most. I am using my Hyper Detailed Illustration LoRA at low weight since that can seem to help.

  • @oskar4239
    @oskar4239 2 месяца назад +1

    This is awesome!

  • @TripHanoi
    @TripHanoi 8 дней назад +1

    It is too perfect

  • @gumvue.studio
    @gumvue.studio 2 месяца назад +1

    thanks, this is very helpfull

    • @KLEEBZTECH
      @KLEEBZTECH  2 месяца назад +1

      You're welcome! I am not sure how useful but I figure the more info people have the better we can all learn.

  • @MichauxJHyatt
    @MichauxJHyatt 2 месяца назад +1

    Pretty dope!

  • @travislittle4381
    @travislittle4381 22 дня назад

    This is a nice walkthrough. I'm still curious what your training images looked like to get this concept, and how you prepared them. Were they standard-sized people superimposed onto close-up objects? Did you use IMG2IMG to reimagine photos of actual miniatures? (I did this with "Precious Memories" figurines, to some very humorous effects). Tell me/us more what the prep was BEFORE Gym Class.

    • @KLEEBZTECH
      @KLEEBZTECH  22 дня назад +1

      There were a mix of things. Some from movie stills and others from images made with photoshop to appear as miniature people. Then several trainings and using those images to get the final.

  • @mauricioc.almeida2482
    @mauricioc.almeida2482 2 месяца назад

    Thank you very much for your Stable Diffusion tutorials. They're always great! Is it possible to install Fooocus on my local machine?

    • @KLEEBZTECH
      @KLEEBZTECH  2 месяца назад +1

      @@mauricioc.almeida2482 older video but install method is the same. ruclips.net/video/j1WuQndmgFE/видео.html

    • @mauricioc.almeida2482
      @mauricioc.almeida2482 2 месяца назад +1

      @@KLEEBZTECH Hello, thank you for your response. I may have asked the question incorrectly. Let me try again: is it possible to install the miniature (safetensors) model in Fooocus 2.5.5? I already have Fooocus installed locally. I tried with the current version (Fooocus 2.5.5) and the result was people with plastic doll characteristics. I'm running the checkpoint "juggernautXL_v8Rundiffusion" in Fooocus.

    • @KLEEBZTECH
      @KLEEBZTECH  2 месяца назад +1

      No. I have not released an SDXL version. I have tried and will attempt again in the near future but so far the results have always been terrible.

  • @TripHanoi
    @TripHanoi 8 дней назад +1

    Good

  • @CyberwizardProductions
    @CyberwizardProductions 2 месяца назад +1

    i would have 1. used photoshop to scale people down and position them on objects, then used those images for the data set and 2. used a unique trigger word that wasn't likely to be in the base model's training set.

    • @KLEEBZTECH
      @KLEEBZTECH  2 месяца назад

      Tried a unique trigger. Made no difference in the end. I thought of doing the images of people and scaling down but just didn't feel I would get them looking right. I am not expert on that.

    • @KLEEBZTECH
      @KLEEBZTECH  2 месяца назад

      Plus Flux knows what miniature people are but it makes them plastic so figured this would just replace that look.

  • @deadpool_gamer8948
    @deadpool_gamer8948 Месяц назад

    In Foocus
    How do I use the program with the CPU?

  • @alpenjager3986
    @alpenjager3986 Месяц назад

    Where have you gone? I haven't seen your content related to Text to Image generator for a long time.

    • @KLEEBZTECH
      @KLEEBZTECH  Месяц назад

      New videos soon. Things outside of RUclips taking my time up recently.

  • @Palestine.0
    @Palestine.0 6 дней назад +1

    why u don't upload any video we miss your video

    • @KLEEBZTECH
      @KLEEBZTECH  4 дня назад

      Soon. Stuff outside of this has taken my time.

  • @kayanaclarasasti8973
    @kayanaclarasasti8973 5 дней назад

    I try this lora but the skin like a plastic not natural, do you know how to make it like natural skin, i try it on comfyui

    • @KLEEBZTECH
      @KLEEBZTECH  5 дней назад

      @@kayanaclarasasti8973 you can use things like realism Loras. Most of the issue is Flux itself and also how I think it is working and just making the normal plastic figures it would create with much more realistic ones but not quite all the way. I have two versions but each is better for certain things so I try both sometimes.

  • @metanulski
    @metanulski 2 месяца назад +1

    Some Random Tips: Let ChatGPT writ a python program for vs code that does screen captures every x seconds from a video. Tell ChatGPT that you also want to specify the start and end time and that the filename has to include the video name and the timestamp. Then you can start with some screenshots from the full video and go info details based on the result. No need for resolve an going though 100000 captures ;-)