SD15BaseModel

class SD15BaseModel(cfg_scale: float = 0.0, min_cfg_scale: float | None = None, max_cfg_scale: float | None = None, prompts: list[str] | None = None, device: device | None = None)[source]

Bases: StableDiffusionBaseModel

Pre-trained 512x512 Stable Diffusion 1.5.

Parameters:

  • cfg_scale (float) – Classifier-free guidance scale to use during sampling. If 0.0, no CFG is used.

  • min_cfg_scale (float, optional) – If provided, the minimum CFG scale to use during sampling. If provided, this is used together with max_cfg_scale to sample a random CFG scale for each sample.

  • max_cfg_scale (float, optional) – If provided, the maximum CFG scale to use during sampling. If provided, this is used together with min_cfg_scale to sample a random CFG scale for each sample.

  • prompts (list[str], optional) – List of prompts to use for sampling if no prompt is provided through the input. If None, a default set of prompts is used.

  • device (torch.device, optional) – Device to load the model on. If None, uses the default.