Controlnet Models Selector v2

Documentation

• Class name: SeargeControlnetModels
• Category: Searge/UI/Inputs
• Output node: False

The SeargeControlnetModels node is designed to facilitate the selection and application of various controlnet models within a generative AI pipeline. It abstracts the complexity of choosing and integrating different controlnet models, which are essential for modifying or enhancing generated content based on specific control parameters or conditions.

Input types

Required

• clip_vision
  • Specifies the CLIP vision model to be used, potentially including a 'none' option to indicate no specific CLIP vision model is required. This input is crucial for determining the visual understanding context for the controlnet model.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str
• canny_checkpoint
  • Refers to the specific checkpoint for the Canny edge detection model within the controlnet framework, including a 'none' option. It's used to apply edge detection features to the generated content.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str
• depth_checkpoint
  • Indicates the checkpoint for the depth estimation model, allowing for depth-aware modifications to the generated content. This parameter includes a 'none' option to skip depth processing.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str
• recolor_checkpoint
  • Specifies the checkpoint for the recoloring model, enabling color adjustments or transformations in the generated content. Includes a 'none' option for cases where recoloring is not required.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str
• sketch_checkpoint
  • Denotes the checkpoint for the sketch model, used to apply sketch-like effects or transformations to the generated content. A 'none' option is included for flexibility.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str
• custom_checkpoint
  • Allows for the specification of a custom controlnet model checkpoint, providing additional flexibility in content modification. Includes a 'none' option.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str

Optional

• data
  • Optional stream of data that can be used for additional processing or as part of the controlnet model's input. This parameter provides flexibility in handling complex workflows.
  • Comfy dtype: SRG_DATA_STREAM
  • Python dtype: str

Output types

• data
  • Comfy dtype: SRG_DATA_STREAM
  • The output data stream resulting from the application of the controlnet models, encapsulating all modifications and enhancements made to the generated content.
  • Python dtype: str

Usage tips

• Infra type: CPU
• Common nodes: unknown

Source code

class SeargeControlnetModels:
    @classmethod
    def INPUT_TYPES(cls):
        return {
            "required": {
                "clip_vision": (UI.CLIP_VISION_WITH_NONE(),),
                "canny_checkpoint": (UI.CONTROLNETS_WITH_NONE(),),
                "depth_checkpoint": (UI.CONTROLNETS_WITH_NONE(),),
                "recolor_checkpoint": (UI.CONTROLNETS_WITH_NONE(),),
                "sketch_checkpoint": (UI.CONTROLNETS_WITH_NONE(),),
                "custom_checkpoint": (UI.CONTROLNETS_WITH_NONE(),),
            },
            "optional": {
                "data": ("SRG_DATA_STREAM",),
            },
        }

    RETURN_TYPES = ("SRG_DATA_STREAM",)
    RETURN_NAMES = ("data",)
    FUNCTION = "get"

    CATEGORY = UI.CATEGORY_UI_INPUTS

    @staticmethod
    def create_dict(clip_vision, canny_checkpoint, depth_checkpoint, recolor_checkpoint, sketch_checkpoint,
                    custom_checkpoint):
        return {
            UI.F_CLIP_VISION_CHECKPOINT: clip_vision,
            UI.F_CANNY_CHECKPOINT: canny_checkpoint,
            UI.F_DEPTH_CHECKPOINT: depth_checkpoint,
            UI.F_RECOLOR_CHECKPOINT: recolor_checkpoint,
            UI.F_SKETCH_CHECKPOINT: sketch_checkpoint,
            UI.F_CUSTOM_CHECKPOINT: custom_checkpoint,
        }

    def get(self, clip_vision, canny_checkpoint, depth_checkpoint, recolor_checkpoint, sketch_checkpoint,
            custom_checkpoint, data=None):
        if data is None:
            data = {}

        data[UI.S_CONTROLNET_MODELS] = self.create_dict(
            clip_vision,
            canny_checkpoint,
            depth_checkpoint,
            recolor_checkpoint,
            sketch_checkpoint,
            custom_checkpoint,
        )

        return (data,)