Load Advanced ControlNet Model (diff) 🛂🅐🅒🅝

Documentation

Class name: DiffControlNetLoaderAdvanced
Category: Adv-ControlNet 🛂🅐🅒🅝
Output node: False

The DiffControlNetLoaderAdvanced node is designed for loading advanced control net models with specific modifications or differences. It focuses on enhancing the control net's capabilities by allowing the integration of additional parameters and customizations, thereby enabling more sophisticated control over the model's behavior.

Input types

Required

model
- Specifies the model to be used in conjunction with the control net, allowing for a tailored approach to loading and applying control nets.
- Comfy dtype: MODEL
- Python dtype: str
control_net_name
- The name of the control net to be loaded. This parameter identifies the specific control net file within a predefined directory structure.
- Comfy dtype: COMBO[STRING]
- Python dtype: str

Optional

timestep_keyframe
- An optional parameter that allows for the specification of timestep keyframes, providing temporal control over the application of the control net.
- Comfy dtype: TIMESTEP_KEYFRAME
- Python dtype: TimestepKeyframeGroup

Output types

control_net
- Comfy dtype: CONTROL_NET
- Returns the loaded control net, enhanced with the specified model and optional timestep keyframes for advanced control.
- Python dtype: ControlNet

Usage tips

Infra type: CPU
Common nodes: unknown

Source code

class DiffControlNetLoaderAdvanced:
    @classmethod
    def INPUT_TYPES(s):
        return {
            "required": {
                "model": ("MODEL",),
                "control_net_name": (folder_paths.get_filename_list("controlnet"), )
            },
            "optional": {
                "timestep_keyframe": ("TIMESTEP_KEYFRAME", ),
            }
        }
    
    RETURN_TYPES = ("CONTROL_NET", )
    FUNCTION = "load_controlnet"

    CATEGORY = "Adv-ControlNet 🛂🅐🅒🅝"

    def load_controlnet(self, control_net_name, model,
                        timestep_keyframe: TimestepKeyframeGroup=None
                        ):
        controlnet_path = folder_paths.get_full_path("controlnet", control_net_name)
        controlnet = load_controlnet(controlnet_path, timestep_keyframe, model)
        if is_advanced_controlnet(controlnet):
            controlnet.verify_all_weights()
        return (controlnet,)

2.5 KiB Raw Blame History