SD 3.5 Medium Anime

Viewing 2 reply threads
  • Author
    Posts
    • January 2, 2025 at 2:27 am #16575
      Yuē bīn
      Participant

        Anyone already attempted at this? Especially in Loras/Checkpoint. I like the prompt adherence better than SDXL in terms of dynamic actions (prefer it over pure aesthetic of models like Animagine, Illustration SDXL) But the issue is i dont have the resources to fine tuned a SD 3.5 Medium into an anime model

      • January 2, 2025 at 7:42 pm #16593
        AndrewAndrew
        Keymaster

          I haven’t trained any but kohya supports training SD 3.5 in the sd3 branch.

          https://github.com/kohya-ss/sd-scripts/tree/sd3

        • January 7, 2025 at 10:34 pm #16736
          Yuē bīn
          Participant

            I tried with ai-toolkit, i got bad results lol I’ll probably wait for fine tune models then ill work on top of it. SDXL anime models doesn’t have prompt adherence I need, they mostly focus on aesthetic

        • Author
          Posts
        Viewing 2 reply threads
        • You must be logged in to reply to this topic.