ComfyUI/comfy_extras/nodes_apg.py

import torch
from typing_extensions import override

from comfy_api.latest import ComfyExtension, io


def project(v0, v1):
    v1 = torch.nn.functional.normalize(v1, dim=[-1, -2, -3])
    v0_parallel = (v0 * v1).sum(dim=[-1, -2, -3], keepdim=True) * v1
    v0_orthogonal = v0 - v0_parallel
    return v0_parallel, v0_orthogonal

class APG(io.ComfyNode):
    @classmethod
    def define_schema(cls) -> io.Schema:
        return io.Schema(
            node_id="APG",
            display_name="Adaptive Projected Guidance",
            category="sampling/custom_sampling",
            description="Applies Adaptive Projected Guidance to a model, decomposing CFG guidance into parallel and orthogonal components with optional momentum and norm thresholding for improved sampling quality.",
            short_description="Decomposes CFG guidance with projection and normalization.",
            inputs=[
                io.Model.Input("model"),
                io.Float.Input(
                    "eta",
                    default=1.0,
                    min=-10.0,
                    max=10.0,
                    step=0.01,
                    tooltip="Controls the scale of the parallel guidance vector. Default CFG behavior at a setting of 1.",
                ),
                io.Float.Input(
                    "norm_threshold",
                    default=5.0,
                    min=0.0,
                    max=50.0,
                    step=0.1,
                    tooltip="Normalize guidance vector to this value, normalization disable at a setting of 0.",
                ),
                io.Float.Input(
                    "momentum",
                    default=0.0,
                    min=-5.0,
                    max=1.0,
                    step=0.01,
                    tooltip="Controls a running average of guidance during diffusion, disabled at a setting of 0.",
                ),
            ],
            outputs=[io.Model.Output()],
        )

    @classmethod
    def execute(cls, model, eta, norm_threshold, momentum) -> io.NodeOutput:
        running_avg = 0
        prev_sigma = None

        def pre_cfg_function(args):
            nonlocal running_avg, prev_sigma

            if len(args["conds_out"]) == 1:
                return args["conds_out"]

            cond = args["conds_out"][0]
            uncond = args["conds_out"][1]
            sigma = args["sigma"][0]
            cond_scale = args["cond_scale"]

            if prev_sigma is not None and sigma > prev_sigma:
                running_avg = 0
            prev_sigma = sigma

            guidance = cond - uncond

            if momentum != 0:
                if not torch.is_tensor(running_avg):
                    running_avg = guidance
                else:
                    running_avg = momentum * running_avg + guidance
                guidance = running_avg

            if norm_threshold > 0:
                guidance_norm = guidance.norm(p=2, dim=[-1, -2, -3], keepdim=True)
                scale = torch.minimum(
                    torch.ones_like(guidance_norm),
                    norm_threshold / guidance_norm
                )
                guidance = guidance * scale

            guidance_parallel, guidance_orthogonal = project(guidance, cond)
            modified_guidance = guidance_orthogonal + eta * guidance_parallel

            modified_cond = (uncond + modified_guidance) + (cond - uncond) / cond_scale

            return [modified_cond, uncond] + args["conds_out"][2:]

        m = model.clone()
        m.set_model_sampler_pre_cfg_function(pre_cfg_function)
        return io.NodeOutput(m)


class ApgExtension(ComfyExtension):
    @override
    async def get_node_list(self) -> list[type[io.ComfyNode]]:
        return [
            APG,
        ]

async def comfy_entrypoint() -> ApgExtension:
    return ApgExtension()