[Python] How to run Stable Diffusion 3.5 Large on 4090?

Discussão em 'Python' iniciado por Stack, Junho 17, 2025.

  1. Stack

    Stack Membro Participativo

    I have 4090 and this code:

    login(HF_TOKEN)
    self.pipe = StableDiffusion3Pipeline.from_pretrained("stabilityai/stable-diffusion-3.5-large", torch_dtype=torch.bfloat16)

    self.pipe = self.pipe.to("cuda")


    This fails to load to vram due to oom. What is the correct way to run it?

    Tried using other floats but did not work

    Continue reading...

Compartilhe esta Página