# Documentation - Class name: LoadInstanceFusersNode - Category: instance/loaders - Output node: False 此节点旨在从指定目录加载和准备实例融合器模型,并根据提供的参数调整其比例。它促进了融合器模型在实例扩散过程中的动态整合,从而实现对实例特定特征融合的增强控制。 # Input types ## Required - model_filename - 指定要加载的模型文件名。这个参数对于从指定目录中识别和检索正确的融合器模型至关重要。 - Comfy dtype: COMBO[STRING] - Python dtype: str - fusers_scale - 确定要应用于融合器的缩放因子,影响它们在实例融合过程中的作用。这允许对融合器对生成实例的影响进行微调。 - Comfy dtype: FLOAT - Python dtype: float # Output types - fusers - 提供已加载和缩放的融合器模型列表,可以直接集成到实例扩散过程中。 - Comfy dtype: FUSERS - Python dtype: Dict[str, List[torch.nn.Module]] ## Usage tips - Infra type: `GPU` - Common nodes: unknown ## Source code ```python class LoadInstanceFusersNode: @classmethod def INPUT_TYPES(s): return {"required": { "model_filename": (get_model_list(constants.INSTANCE_FUSERS_DIR),), "fusers_scale": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01}), }} RETURN_TYPES = ("FUSERS",) FUNCTION = "load_model" CATEGORY = "instance/loaders" def load_model(self, model_filename: str, fusers_scale: float): checkpoint = load_checkpoint( constants.INSTANCE_FUSERS_DIR, model_filename) fusers_list = prepare_fusers(checkpoint, fusers_scale) fusers = { 'model_list': fusers_list } return (fusers,) ```