secondary_diffusion.py

from dataclasses import dataclass
import math
from functools import partial

import torch
import torch.nn as nn


@dataclass
class DiffusionOutput:
    v: torch.Tensor
    pred: torch.Tensor
    eps: torch.Tensor
    
    
class ConvBlock(nn.Sequential):
    def __init__(self, c_in, c_out):
        super().__init__(
            nn.Conv2d(c_in, c_out, 3, padding=1),
            nn.ReLU(inplace=True),
        )


class SkipBlock(nn.Module):
    def __init__(self, main, skip=None):
        super().__init__()
        self.main = nn.Sequential(*main)
        self.skip = skip if skip else nn.Identity()

    def forward(self, input):
        return torch.cat([self.main(input), self.skip(input)], dim=1)


class FourierFeatures(nn.Module):
    def __init__(self, in_features, out_features, std=1.):
        super().__init__()
        assert out_features % 2 == 0
        self.weight = nn.Parameter(torch.randn([out_features // 2, in_features]) * std)

    def forward(self, input):
        f = 2 * math.pi * input @ self.weight.T
        return torch.cat([f.cos(), f.sin()], dim=-1)


class SecondaryDiffusionImageNet(nn.Module):
    def __init__(self):
        super().__init__()
        c = 64  # The base channel count

        self.timestep_embed = FourierFeatures(1, 16)

        self.net = nn.Sequential(
            ConvBlock(3 + 16, c),
            ConvBlock(c, c),
            SkipBlock([
                nn.AvgPool2d(2),
                ConvBlock(c, c * 2),
                ConvBlock(c * 2, c * 2),
                SkipBlock([
                    nn.AvgPool2d(2),
                    ConvBlock(c * 2, c * 4),
                    ConvBlock(c * 4, c * 4),
                    SkipBlock([
                        nn.AvgPool2d(2),
                        ConvBlock(c * 4, c * 8),
                        ConvBlock(c * 8, c * 4),
                        nn.Upsample(scale_factor=2, mode='bilinear', align_corners=False),
                    ]),
                    ConvBlock(c * 8, c * 4),
                    ConvBlock(c * 4, c * 2),
                    nn.Upsample(scale_factor=2, mode='bilinear', align_corners=False),
                ]),
                ConvBlock(c * 4, c * 2),
                ConvBlock(c * 2, c),
                nn.Upsample(scale_factor=2, mode='bilinear', align_corners=False),
            ]),
            ConvBlock(c * 2, c),
            nn.Conv2d(c, 3, 3, padding=1),
        )

    def forward(self, input, t):
        timestep_embed = expand_to_planes(self.timestep_embed(t[:, None]), input.shape)
        v = self.net(torch.cat([input, timestep_embed], dim=1))
        alphas, sigmas = map(partial(append_dims, n=v.ndim), t_to_alpha_sigma(t))
        pred = input * alphas - v * sigmas
        eps = input * sigmas + v * alphas
        return DiffusionOutput(v, pred, eps)


class SecondaryDiffusionImageNet2(nn.Module):
    def __init__(self):
        super().__init__()
        c = 64  # The base channel count
        cs = [c, c * 2, c * 2, c * 4, c * 4, c * 8]

        self.timestep_embed = FourierFeatures(1, 16)
        self.down = nn.AvgPool2d(2)
        self.up = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=False)

        self.net = nn.Sequential(
            ConvBlock(3 + 16, cs[0]),
            ConvBlock(cs[0], cs[0]),
            SkipBlock([
                self.down,
                ConvBlock(cs[0], cs[1]),
                ConvBlock(cs[1], cs[1]),
                SkipBlock([
                    self.down,
                    ConvBlock(cs[1], cs[2]),
                    ConvBlock(cs[2], cs[2]),
                    SkipBlock([
                        self.down,
                        ConvBlock(cs[2], cs[3]),
                        ConvBlock(cs[3], cs[3]),
                        SkipBlock([
                            self.down,
                            ConvBlock(cs[3], cs[4]),
                            ConvBlock(cs[4], cs[4]),
                            SkipBlock([
                                self.down,
                                ConvBlock(cs[4], cs[5]),
                                ConvBlock(cs[5], cs[5]),
                                ConvBlock(cs[5], cs[5]),
                                ConvBlock(cs[5], cs[4]),
                                self.up,
                            ]),
                            ConvBlock(cs[4] * 2, cs[4]),
                            ConvBlock(cs[4], cs[3]),
                            self.up,
                        ]),
                        ConvBlock(cs[3] * 2, cs[3]),
                        ConvBlock(cs[3], cs[2]),
                        self.up,
                    ]),
                    ConvBlock(cs[2] * 2, cs[2]),
                    ConvBlock(cs[2], cs[1]),
                    self.up,
                ]),
                ConvBlock(cs[1] * 2, cs[1]),
                ConvBlock(cs[1], cs[0]),
                self.up,
            ]),
            ConvBlock(cs[0] * 2, cs[0]),
            nn.Conv2d(cs[0], 3, 3, padding=1),
        )

    def forward(self, input, t):
        timestep_embed = expand_to_planes(self.timestep_embed(t[:, None]), input.shape)
        v = self.net(torch.cat([input, timestep_embed], dim=1))
        alphas, sigmas = map(partial(append_dims, n=v.ndim), t_to_alpha_sigma(t))
        pred = input * alphas - v * sigmas
        eps = input * sigmas + v * alphas
        return DiffusionOutput(v, pred, eps)
    

def append_dims(x, n):
    return x[(Ellipsis, *(None,) * (n - x.ndim))]


def expand_to_planes(x, shape):
    return append_dims(x, len(shape)).repeat([1, 1, *shape[2:]])


def t_to_alpha_sigma(t):
    return torch.cos(t * math.pi / 2), torch.sin(t * math.pi / 2)