10x Faster Images With This SIMPLE TRICK

Поділитися
Вставка
  • Опубліковано 20 вер 2024

КОМЕНТАРІ • 18

  • @NextDiffusion
    @NextDiffusion  10 місяців тому +2

    SPEED🚀 Happy Generating Everyone!
    Written Tutorial: www.nextdiffusion.ai/tutorials/how-to-speed-up-image-generation-with-lcm-loras-a1111

  • @jag9022
    @jag9022 10 місяців тому +3

    Underrated channel keep it up with your tutorials!

    • @NextDiffusion
      @NextDiffusion  10 місяців тому

      Thanks, will do!

    • @Navienzhal
      @Navienzhal 10 місяців тому

      @@NextDiffusion i think yt algorithms purposefully not advertise any videos with ai theme so there is a little chance to get high humber of subscribers
      great summary anyway

  • @user-je7qy5ey3y
    @user-je7qy5ey3y 10 місяців тому +1

    It's like you make a cocktail of Laura's work together to make a fantastic job

  • @robroufla
    @robroufla 6 місяців тому

    Only LCM and LCM Karras samplers give good results from my XYZ tests with 8 steps and it's strangely much slower than without the lcm lora

  • @yoyo1poe
    @yoyo1poe 10 місяців тому +1

    I tried the LCM extension, it's lightning fast but the model is very limited. I didn't know there existed Loras though. I need to try these

  • @user-je7qy5ey3y
    @user-je7qy5ey3y 10 місяців тому +1

    I can't believe how much is fast😀👍

  • @JLITZ88
    @JLITZ88 10 місяців тому

    How do you combine LCM with animation in deforum?

    • @NextDiffusion
      @NextDiffusion  10 місяців тому

      put the lora in the prompt in deforum and adjust settings accordingly.

    • @JLITZ88
      @JLITZ88 10 місяців тому

      ​@@NextDiffusion Hi, so i tried pasting into the deforum prompt but see no increase in s/it have you tried it out? the lora model is included in the deforum model section working on it

  • @monkeymediapl
    @monkeymediapl 9 місяців тому

    Apart from the topic of video, did you test ComfyUi workflow? Why you choose A1111? Which one is better and why? Will be very greatful for answer.

  • @TheSslawek89
    @TheSslawek89 10 місяців тому

    How to make it work with SDXL models

    • @NextDiffusion
      @NextDiffusion  10 місяців тому

      Same method as explained in the video, but instead of the SD 1.5 version you need to download the LCM SD XL LoRA. Link provided in description.

  • @user-je7qy5ey3y
    @user-je7qy5ey3y 10 місяців тому

    With this you can create animation with this file

  • @nermal93
    @nermal93 10 місяців тому

    Is this working with img2img?

    • @NextDiffusion
      @NextDiffusion  10 місяців тому +2

      Certainly, I overlooked mentioning the LCM LoRA's can be applied to the following tasks:
      - Image-to-image
      - Inpainting
      - ControlNet

  • @Thex-W.I.T.C.H.-xMaster
    @Thex-W.I.T.C.H.-xMaster 10 місяців тому

    Skynet will be here within the next 5 to 7 years. Luckily the planet is going to explode 🌎 💥... moments after the machines become selfaware 🥴🤞. Unfortunately for them 😕.