How To Install Stable Diffusion Models: InvokeAI

Поділитися
Вставка
  • Опубліковано 17 лис 2024

КОМЕНТАРІ • 36

  • @CC-so7zh
    @CC-so7zh Рік тому +7

    thanks man, most tutorials are for people who know what they are doing. this video is a lot helpful for a noob like me

  • @whiskerwalks
    @whiskerwalks Місяць тому

    Having trouble install model but maybe cause my interface looks different with invoke 3.2.0 😑

  • @RazielDark
    @RazielDark Рік тому +2

    Thanks for pointing out invokeAI it's really good!

  • @chapli_kabbabgaming9812
    @chapli_kabbabgaming9812 Рік тому

    sir i have a question .
    i am using stable diffusion i generating a model but not making properly image face eyes nose making ugly 90% ugly in each image so how i fix it please help me
    thank you

  • @akas1433
    @akas1433 Рік тому

    I don't know if i'm the only one , but when I wan't to load a model , it say KeyError: 'time_embed.0.weight', any clues ?

  • @MrVitaliyAT
    @MrVitaliyAT Рік тому

    i have lora, how I can use it?

  • @jeanies1090
    @jeanies1090 Рік тому +1

    Does it work if I use a Mac?

    • @allyourtechai
      @allyourtechai  Рік тому

      You can install invoke ai on an m1 or m2 Mac, it’s just very slow.

  • @granddude1242
    @granddude1242 Рік тому

    I'm trying to get started, getting confused! Should I use my Tab 8+ 12Gigs, 512, or Chromebook, or Windows 11 Gamer that I've put up cause I don't use? I'm 75 yrs old. Systems are getting confusing.With so many.Dp\o I have to use Discord to operate

  • @digiministrator
    @digiministrator Рік тому

    Hello,
    Is it possible to add Kandinsky 2.2 as a model in InvokeAI?

    • @digiministrator
      @digiministrator Рік тому

      :) RTX 3050 4GB, It didn't work for me, hardware is too expensive for software, I don't understand why they don't lighten it.

  • @phoneflasher
    @phoneflasher Рік тому

    its working off my CPU how do I get it to use my GPU?
    AMD 7900x CPU,
    AMD 7900xt 20GB GPU

  • @giuseppedaizzole7025
    @giuseppedaizzole7025 10 місяців тому

    Hi there...very good video....i have a question....is this possible in win 7?...thanks

    • @allyourtechai
      @allyourtechai  10 місяців тому +1

      It looks like it may be possible sketchbooky.wordpress.com/2023/11/20/how-to-install-invokeai-on-windows-7-sp1/

    • @giuseppedaizzole7025
      @giuseppedaizzole7025 10 місяців тому

      Thanks alot for the link, really appreciate.@@allyourtechai

  • @m12652
    @m12652 8 місяців тому

    2:07 as of today, never used it before, theres something wrong with the site. Models page just lists images and clicking one just takes you to image, no option to download a model. Very stupid site considering the context... off to hugging face.

  • @swiftals
    @swiftals Рік тому +1

    Very nice video and well explained . May I ask you which gpu are you using ?

    • @allyourtechai
      @allyourtechai  Рік тому +1

      Thank you! I'm using an Nvidia RTX 3090. It's great for AI workloads because of the 24GB of VRAM. If I were serious about getting into this stuff I would try to get a card with a minimum of 12GB of VRAM (and from nvidia since AMD doesn't have tensor cores for accelerating these workloads)

    • @swiftals
      @swiftals Рік тому +1

      Thank you . I'm using gtx 1070 with 8gb and is very slow like 2 minutes per image , also on m1 macbook air with 16gb of ram it takes around 35 to 40 sec with 20 steps and 512x512 pixels but I get on swap where I don't like it . Maybe I should upgrade to rtx 3060 or rtx 4070 or Intel arc a770 16gb.

    • @allyourtechai
      @allyourtechai  Рік тому +1

      @@swiftals Any of the RTX cards should do a significantly better job. FOr the price, the 4070 looks pretty decent

  • @Nawaf511-t1z
    @Nawaf511-t1z Рік тому +1

    Thx👍 I don't know how to download openjourney-v4 😂

  • @Nicocotier12
    @Nicocotier12 Рік тому

    Thanks for the video. Is it the same thing for add on ? I see a lot in the community a lot of people using controlnet and other thing they add to stable diffusion etc.
    If you understand those thing it would be a great help to explain what those are

    • @allyourtechai
      @allyourtechai  Рік тому

      I have an entire video dedicated to controlnet. How to install it in automatic1111 and how to use it. Check it out!

  • @joestradamus75
    @joestradamus75 Рік тому

    when I render an image, the character's in the picture are hardly ever centered. It's either zoomed in on the face to much or if I have 2 people in a scene, one person may be off to the side just barely in view. Can you give instructions on how to get everyone in a picture centered..etc...

    • @allyourtechai
      @allyourtechai  Рік тому +2

      YOu can typically guide the image generation similar to how you would a real photo. Something like this for example:
      surprised photo of XYZ person taken by professional photographer, 35mm, F/2.8, 4k, photojournalism, busy background, volumetric lighting, sharp focus. I use chatgpt to help with my prompts quite a bit. YOu might also find it useful: ua-cam.com/video/qMvHQc2QzmU/v-deo.html

    • @joestradamus75
      @joestradamus75 Рік тому

      @@allyourtechai thankyou, I'll try that.

  • @BSDOWNZ
    @BSDOWNZ Рік тому +1

    Cheers

  • @HalitHarun
    @HalitHarun Рік тому

    get hold of this man

  • @ClocklessHours
    @ClocklessHours 6 місяців тому

    you look ai generated and ai animated

  • @14qw4xvg1q4
    @14qw4xvg1q4 Рік тому +1

    are you an AI?