Perp-Neg (DEPRECATED by PerpNegGuider)

Documentation

• Class name: PerpNeg
• Category: _for_testing
• Output node: False

The PerpNeg node is designed to adjust the conditioning of a model by applying a perpendicular negative guidance technique. This technique modifies the input conditioning vectors to enhance the generation of content that diverges from undesired directions, effectively guiding the model towards more desirable outputs.

Input types

Required

• model
  • The model parameter represents the generative model to which the perpendicular negative guidance will be applied. It is crucial for determining the base behavior and capabilities of the node.
  • Comfy dtype: MODEL
  • Python dtype: comfy.model_management.MODEL
• empty_conditioning
  • The empty_conditioning parameter is used to provide a baseline or neutral conditioning context, against which positive and negative deviations are calculated and adjusted.
  • Comfy dtype: CONDITIONING
  • Python dtype: CONDITIONING
• neg_scale
  • The neg_scale parameter controls the intensity of the negative guidance applied, allowing for fine-tuning of the perpendicular negative effect on the model's output.
  • Comfy dtype: FLOAT
  • Python dtype: float

Output types

• model
  • Comfy dtype: MODEL
  • The output is a modified version of the input model, adjusted with perpendicular negative guidance to influence its generation process towards desired outcomes.
  • Python dtype: comfy.model_management.MODEL

Usage tips

• Infra type: CPU
• Common nodes: unknown

Source code

class PerpNeg:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {"model": ("MODEL", ),
                             "empty_conditioning": ("CONDITIONING", ),
                             "neg_scale": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 100.0, "step": 0.01}),
                            }}
    RETURN_TYPES = ("MODEL",)
    FUNCTION = "patch"

    CATEGORY = "_for_testing"

    def patch(self, model, empty_conditioning, neg_scale):
        m = model.clone()
        nocond = comfy.sampler_helpers.convert_cond(empty_conditioning)

        def cfg_function(args):
            model = args["model"]
            noise_pred_pos = args["cond_denoised"]
            noise_pred_neg = args["uncond_denoised"]
            cond_scale = args["cond_scale"]
            x = args["input"]
            sigma = args["sigma"]
            model_options = args["model_options"]
            nocond_processed = comfy.samplers.encode_model_conds(model.extra_conds, nocond, x, x.device, "negative")

            (noise_pred_nocond,) = comfy.samplers.calc_cond_batch(model, [nocond_processed], x, sigma, model_options)

            cfg_result = x - perp_neg(x, noise_pred_pos, noise_pred_neg, noise_pred_nocond, neg_scale, cond_scale)
            return cfg_result

        m.set_model_sampler_cfg_function(cfg_function)

        return (m, )