TwoSamplersForMask Upscaler Provider

Documentation

• Class name: TwoSamplersForMaskUpscalerProvider
• Category: ImpactPack/Upscale
• Output node: False

This node provides a mechanism to upscale images using two distinct samplers, each responsible for different regions of the image, based on a mask. It enables the selective application of upscaling techniques to enhance image quality or add details in a controlled manner.

Input types

Required

• scale_method
  • Specifies the method used for scaling during the upscaling process.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str
• full_sample_schedule
  • Defines the sampling schedule for the full image upscaling.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: Dict[str, Any]
• use_tiled_vae
  • Indicates whether a tiled VAE approach is used for upscaling.
  • Comfy dtype: BOOLEAN
  • Python dtype: bool
• base_sampler
  • The primary sampler used for upscaling the image, applied to the regions not covered by the mask.
  • Comfy dtype: KSAMPLER
  • Python dtype: torch.nn.Module
• mask_sampler
  • A specialized sampler used for upscaling the regions of the image specified by the mask.
  • Comfy dtype: KSAMPLER
  • Python dtype: torch.nn.Module
• mask
  • A binary mask indicating the regions of the image to be specifically upscaled by the mask_sampler.
  • Comfy dtype: MASK
  • Python dtype: torch.Tensor
• vae
  • The variational autoencoder used in the upscaling process.
  • Comfy dtype: VAE
  • Python dtype: torch.nn.Module
• tile_size
  • Specifies the size of the tiles when using a tiled VAE approach.
  • Comfy dtype: INT
  • Python dtype: int

Optional

• full_sampler_opt
  • Options for the full sampler used in the upscaling process.
  • Comfy dtype: KSAMPLER
  • Python dtype: Dict[str, Any]
• upscale_model_opt
  • Options for the upscale model used in the upscaling process.
  • Comfy dtype: UPSCALE_MODEL
  • Python dtype: Dict[str, Any]
• pk_hook_base_opt
  • Options for the base hook in the upscaling process.
  • Comfy dtype: PK_HOOK
  • Python dtype: Dict[str, Any]
• pk_hook_mask_opt
  • Options for the mask hook in the upscaling process.
  • Comfy dtype: PK_HOOK
  • Python dtype: Dict[str, Any]
• pk_hook_full_opt
  • Options for the full hook in the upscaling process.
  • Comfy dtype: PK_HOOK
  • Python dtype: Dict[str, Any]

Output types

• upscaler
  • Comfy dtype: UPSCALER
  • The upscaled latent image after applying both the base and mask samplers.
  • Python dtype: Dict[str, torch.Tensor]

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class TwoSamplersForMaskUpscalerProvider:
    upscale_methods = ["nearest-exact", "bilinear", "lanczos", "area"]

    @classmethod
    def INPUT_TYPES(s):
        return {"required": {
                     "scale_method": (s.upscale_methods,),
                     "full_sample_schedule": (
                         ["none", "interleave1", "interleave2", "interleave3",
                          "last1", "last2",
                          "interleave1+last1", "interleave2+last1", "interleave3+last1",
                          ],),
                     "use_tiled_vae": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}),
                     "base_sampler": ("KSAMPLER", ),
                     "mask_sampler": ("KSAMPLER", ),
                     "mask": ("MASK", ),
                     "vae": ("VAE",),
                     "tile_size": ("INT", {"default": 512, "min": 320, "max": 4096, "step": 64}),
                     },
                "optional": {
                        "full_sampler_opt": ("KSAMPLER",),
                        "upscale_model_opt": ("UPSCALE_MODEL", ),
                        "pk_hook_base_opt": ("PK_HOOK", ),
                        "pk_hook_mask_opt": ("PK_HOOK", ),
                        "pk_hook_full_opt": ("PK_HOOK", ),
                    }
                }

    RETURN_TYPES = ("UPSCALER", )
    FUNCTION = "doit"

    CATEGORY = "ImpactPack/Upscale"

    def doit(self, scale_method, full_sample_schedule, use_tiled_vae, base_sampler, mask_sampler, mask, vae,
             full_sampler_opt=None, upscale_model_opt=None,
             pk_hook_base_opt=None, pk_hook_mask_opt=None, pk_hook_full_opt=None, tile_size=512):
        upscaler = core.TwoSamplersForMaskUpscaler(scale_method, full_sample_schedule, use_tiled_vae,
                                                   base_sampler, mask_sampler, mask, vae, full_sampler_opt, upscale_model_opt,
                                                   pk_hook_base_opt, pk_hook_mask_opt, pk_hook_full_opt, tile_size=tile_size)
        return (upscaler, )