ComfyUI: Lora? What the F$@K is LORA? Generate anything with Stable Diffusion. and Oops.

แชร์
ฝัง
  • เผยแพร่เมื่อ 7 ก.พ. 2025

ความคิดเห็น • 19

  • @stoiqa
    @stoiqa 10 หลายเดือนก่อน +3

    please do one updated on lora training, I was watching olivios, and unfortunately is dated. straight forward and clear, keep going!!

    • @the-ai-art
      @the-ai-art  10 หลายเดือนก่อน +1

      Thanks for the feedback. I will do one soon.

  • @lagaxxi
    @lagaxxi 2 หลายเดือนก่อน

    you're a genius! thank you so much. I was hours and hours playing with the lora without undertanding hehe

  • @cyberspider78910
    @cyberspider78910 7 หลายเดือนก่อน

    Best part of your tutorial.....You are giving analogus which layman or new bee can understand. e.g. you said "you can consider LORA as an add on to checkpoint with some specific purpose"

    • @icedzinnia
      @icedzinnia 7 หลายเดือนก่อน

      Its n00b3

    • @the-ai-art
      @the-ai-art  7 หลายเดือนก่อน

      Thanks, Glad you find it helpful :)

  • @Mehdi0montahw
    @Mehdi0montahw 9 หลายเดือนก่อน +1

    Wow, Laura is powerful. I didn't know it could be combined with another one. Thank you

    • @the-ai-art
      @the-ai-art  9 หลายเดือนก่อน

      Gladly :)

  • @icedzinnia
    @icedzinnia 7 หลายเดือนก่อน

    Your tutorials are so amazing and valuable. I hope more ppl come across your videos!

    • @the-ai-art
      @the-ai-art  7 หลายเดือนก่อน

      Thank you so much!

  • @kaymifranca
    @kaymifranca 5 หลายเดือนก่อน

    Thank you very much for all your videos, they are helping me a lot to understand this universe

    • @the-ai-art
      @the-ai-art  5 หลายเดือนก่อน

      Happy to hear that!

  • @killbadmashia9225
    @killbadmashia9225 8 หลายเดือนก่อน

    Thank you very much for such a simple explaining, as a beginner I was struggling to understand these basic concepts which I didn't find anyone explained like this.
    I am curious to know if you already have an image , how do you apply loras to that image ?

    • @the-ai-art
      @the-ai-art  7 หลายเดือนก่อน

      In this case you need to pass the image through a VAE encoder to turn it into a latent. then when you plug it into a sampler, denoise it (not too much so you keep the overall details) and use the LORA as you would in the regular workflow.

  • @Oxes
    @Oxes 6 หลายเดือนก่อน

    thanks for this! new sub !

  • @Vikingskog
    @Vikingskog 7 หลายเดือนก่อน

    I have a question, i want to use two lora models that covers roughly the same things, but i want to use a specific prompt for just one of the lora models, and another prompt for the other. For example two different lora models for animals, one realistic and one cartoony, and i want a cat from one and a dog from the other and then have them both in the same image. What would be the best way to do this you think?

    • @the-ai-art
      @the-ai-art  7 หลายเดือนก่อน

      If i understood you correctly, it sounds like you want to merge two images. There are several ways to approach this. It will require use of masks the determine how to merge the two. ImageCompositeMasked is a good place to start. Masking is one of the next videos I plan on making :)

  • @AB-fo3hj
    @AB-fo3hj 5 หลายเดือนก่อน

    Informative, to the point, but the background music is sooo irritating

    • @the-ai-art
      @the-ai-art  4 หลายเดือนก่อน +1

      Yeah, sorry for that.