Generating Images From Text. Stable Diffusion, Explained

Поделиться
HTML-код
  • Опубликовано: 27 окт 2024

Комментарии • 48

  • @feliwein_cc
    @feliwein_cc 2 года назад +13

    man, your explanations are super clear, you have a good abilty to decrease the complexity of a complex topic

    • @underfitted
      @underfitted  2 года назад +1

      I appreciate that!

    • @IxNxI
      @IxNxI Год назад

      I feel like he's always yelling at me lol

  • @teamtom
    @teamtom 2 года назад +14

    this was the best, clearest explanation i have heard so far, thank you!

    • @underfitted
      @underfitted  2 года назад

      Thanks so much for commenting! Really appreciate it!

  • @pr0blem
    @pr0blem Год назад +5

    Oh man, I am data scientist myself, but this explanation is so great and clear. I do not always have time to catch-up with latest ML papers in various areas (who has ;d), but having this clear, high level overview of system architecture with examples is extremely helpful in understanding what are the latest fancy things everyone is talking about. Please do more of this. I'm sensing hundreds of thousands of subscriptions very soon. Great job!

  • @kiric9311
    @kiric9311 2 года назад +3

    clear explanation and easy to understand, it's really helpful for me to understand the Diffusion model!

    • @underfitted
      @underfitted  2 года назад

      Thanks, Kiri! Glad it helped!

  • @kemmounramzy6232
    @kemmounramzy6232 Год назад

    i saw many video about this subject but this the most clear and the best explanation thanks !

  • @ricosrealm
    @ricosrealm Год назад +1

    Great summary and explanation. I'm going to check this out. Thanks!

  • @mgreek31
    @mgreek31 2 года назад +5

    Wow 😲
    This is Awesome Explanation 👍
    I'm up for the task 💪

  • @netanelmad
    @netanelmad 2 года назад +2

    Great video, thanks! Would love to see more on this topic :)

  • @abandonedmuse
    @abandonedmuse 2 года назад +1

    I needed this so badly. Thank you sooooo much! I was looking for something to help me understand what is going on. New sub !

    • @underfitted
      @underfitted  2 года назад +1

      Thanks for subbing!

    • @abandonedmuse
      @abandonedmuse 2 года назад

      @@underfitted of course! How could I not? You are from the same place as me and you love Ai too! I’m just starting to learn. Do you teach at a college or university? Fiu or MDCC? If not you should!

  • @ashwinshetgaonkar6329
    @ashwinshetgaonkar6329 2 года назад +2

    thanks for this concise explaination

  • @khaledabdelhay1543
    @khaledabdelhay1543 9 месяцев назад

    thaaaank you, finally started to get it

  • @sudharsankp
    @sudharsankp 2 года назад +2

    Thanks Santiago, it was very concise. If am not mistaken, i have a suggestion that may be offensive (Am sorry) - Since this is an educational video would be much more easier to intake the content if the video didn't have much of theatrics.

    • @underfitted
      @underfitted  2 года назад

      Thanks for the feedback! Appreciate you taking the time.

  • @shami9221
    @shami9221 Год назад

    man, you're so underrated

  • @ShanilPanara
    @ShanilPanara 2 года назад

    This is awesome bro! Thank you!!

    • @underfitted
      @underfitted  2 года назад

      Sure thing! Thanks for the comment!

  • @edgardodavid4697
    @edgardodavid4697 Год назад +1

    Awesome video.

  • @jespermikkelsen7553
    @jespermikkelsen7553 Год назад

    So cool. Wonder if it's possible to make a compiler using stable diffusion, or translation between languages, human og programming languages....the potential of this technic is mind blowing

  • @IntuitMachine
    @IntuitMachine 2 года назад

    Why is it that the text encoder doesn't need to be retrained with the images of a new diffusion model? Can you explain this part?

    • @underfitted
      @underfitted  2 года назад

      I know we discussed on Twitter about this. Did you ever look into this more?

  • @sonalikumari1335
    @sonalikumari1335 2 года назад +1

    How to find the accuracy of this model ?

    • @underfitted
      @underfitted  2 года назад

      Well, it depends on what you consider to be "accuracy"

  • @sonalikumari1335
    @sonalikumari1335 2 года назад

    You are using stable diffusion module to convert the text to images. Can you please write the code for the steps you explained - noising and reverse denoising. I need it for my project

    • @underfitted
      @underfitted  2 года назад

      Hey Sonali, I'd recommend you look into the original source code of Stable Diffusion. You can also google an example of a Diffusion model and you'll find the code.

  • @abandonedmuse
    @abandonedmuse 2 года назад

    Instead of a gif can we make them .mp4 videos? Can we specify the video encoding?

    • @underfitted
      @underfitted  2 года назад

      Yes you can!

    • @abandonedmuse
      @abandonedmuse 2 года назад

      @@underfitted perfect! i am working on Diffusion Bee at the moment but I still haven’t found a solution for video and this seems like the best idea.

  • @vanedeoz9472
    @vanedeoz9472 5 месяцев назад

    Nice explanation, but the "Nas daily" style of video is so annoying and distracting! I had to mute the video and activate the captions.

  • @joelemons8587
    @joelemons8587 Год назад

    Calm down mate for god's sake

  • @Patapom3
    @Patapom3 2 года назад

    Cool but why are you yelling all the time?