module.py

# pytorch_diffusion + derived encoder decoder
import math
import torch
import torch.nn as nn
import numpy as np


def get_timestep_embedding(timesteps, embedding_dim=256):
    """
    This matches the implementation in Denoising Diffusion Probabilistic Models:
    From Fairseq.
    Build sinusoidal embeddings.
    This matches the implementation in tensor2tensor, but differs slightly
    from the description in Section 3.5 of "Attention Is All You Need".
    """
    assert len(timesteps.shape) == 1

    half_dim = embedding_dim // 2
    emb = math.log(10000) / (half_dim - 1)
    emb = torch.exp(torch.arange(half_dim, dtype=torch.float32) * -emb)
    emb = emb.to(device=timesteps.device)
    emb = timesteps.float()[:, None] * emb[None, :]
    emb = torch.cat([torch.sin(emb), torch.cos(emb)], dim=1)
    if embedding_dim % 2 == 1:  # zero pad
        emb = torch.nn.functional.pad(emb, (0,1,0,0))
    return emb


def nonlinearity(x):
    # swish
    return x*torch.sigmoid(x)


def Normalize(in_channels):
    return torch.nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)

# 把图像转成一维向量
class WeakImgEmb(nn.Module):
    def __init__(self, in_ch, out_ch):
        super(WeakImgEmb, self).__init__()
        self.conv1 = nn.Conv2d(in_ch, out_ch, kernel_size=1,stride=2,padding=0)# 256->128
        self.norm1 = nn.BatchNorm2d(out_ch)
        self.maxpool1 = nn.MaxPool2d(kernel_size=2, stride=2)# 128->64
        self.conv2 = nn.Conv2d(out_ch, out_ch, kernel_size=1,stride=2,padding=0)# 64->32
        self.norm2 = nn.BatchNorm2d(out_ch)
        self.maxpool2 = nn.MaxPool2d(kernel_size=2, stride=2)# 32->16
        self.conv3 = nn.Conv2d(out_ch, out_ch, kernel_size=1,stride=2,padding=0)# 16->8
        
        self.linear = nn.Linear(8*8*3, 256)
    def forward(self, x):
        B,_,_,_ = x.shape
        h = self.norm1(x)
        h = nonlinearity(h)
        h = self.conv1(h)
        mh = self.maxpool1(h)
        
        h = self.norm2(mh)
        h = nonlinearity(h)
        h = self.conv2(h)
        
        h = self.maxpool2(h)
        h_out = self.conv3(h)
        h_out = self.linear(h_out.reshape(B,-1))
        
        return h_out # B,256

class DepthWiseConv(nn.Module):
    def __init__(self, in_ch, out_ch, temb_channels=256):
        super(DepthWiseConv, self).__init__()
        self.conv = nn.Conv2d(in_channels=in_ch, out_channels=out_ch, kernel_size=3,stride=1,padding=1)
        self.depth_conv = nn.Conv2d(in_channels=out_ch, out_channels=out_ch, kernel_size=3, stride=1, padding=1, groups=out_ch)
        self.point_conv = nn.Conv2d(in_channels=out_ch, out_channels=out_ch, kernel_size=1, stride=1,padding=0,groups=1)

        if temb_channels > 0:
            self.temb_proj = torch.nn.Linear(temb_channels,
                                             out_ch)
        # self.relu = nn.ReLU(inplace=False)
        # self.out = nn.Linear(96*img_size*img_size, out_ch)

    def forward(self, x, temb=None):
        h = self.conv(x)

        h = self.depth_conv(h)

        if temb is not None:
            h = h + self.temb_proj(nonlinearity(temb))[:, :, None, None]
        # print("shape of depth conv: {}".format(depth_x.shape))# torch.Size([10, 32, 64, 64])
        h = self.point_conv(h)
        # print("shape of point conv: {}".format(point_x.shape))# torch.Size([10, 96, 64, 64])
        # out = self.out(point_x.reshape(10,-1))

        return h

class SqueezeChannelConv(nn.Module):
    def __init__(self, in_ch, out_ch, se_heads = 6, scale=8):
        super(SqueezeChannelConv, self).__init__()
        self.scale = scale
        self.mid_ch = in_ch // self.scale
        self.num_head = se_heads
        self.se_downs = nn.ModuleList()
        self.fcs = nn.ModuleList()

        for i in range(se_heads):
            self.se_downs.append(nn.Conv2d(in_ch, self.mid_ch, kernel_size=1, stride=1))
            self.se_layer = nn.Sequential(
                nn.AdaptiveAvgPool2d(1),  # Squeeze操作
                nn.Linear(self.mid_ch, self.mid_ch, bias=False),
                nn.SiLU(inplace=True),
                nn.Linear(self.mid_ch, self.mid_ch, bias=False),
                nn.Sigmoid()
            )
            self.fcs.append(self.se_layer)


        self.se_up = nn.Conv2d(self.mid_ch, out_ch,kernel_size=1,stride=1)

    def forward(self, x):
        """
        :param x: B,C,H,W
        :param temp:
        :return:
        """
        b,c,_,_ = x.shape # torch.Size([10, 32, 256, 256])

        se_hidden = 0.0
        for i in range(self.num_head):
            h_se = self.se_downs[i](x) # torch.Size([10, 32 // 8, 256, 256])
            h_w = self.fcs[i][0](h_se).view(b,c//self.scale)
            h_w = self.fcs[i][1:](h_w)
            se_hidden = se_hidden + h_w.view(b,c//self.scale,1,1) * h_se
        # print("shape of se_hiddeb: {}".format(se_hidden.shape))
        se_out = self.se_up(se_hidden)
        return x + se_out





class Upsample(nn.Module):
    def __init__(self, in_channels, with_conv):
        super().__init__()
        self.with_conv = with_conv
        if self.with_conv:
            self.conv = torch.nn.Conv2d(in_channels,
                                        in_channels,
                                        kernel_size=3,
                                        stride=1,
                                        padding=1)

    def forward(self, x):
        x = torch.nn.functional.interpolate(x, scale_factor=2.0, mode="nearest")
        if self.with_conv:
            x = self.conv(x)
        return x


class Downsample(nn.Module):
    def __init__(self, in_channels, with_conv):
        super().__init__()
        self.with_conv = with_conv
        if self.with_conv:
            # no asymmetric padding in torch conv, must do it ourselves
            self.conv = torch.nn.Conv2d(in_channels,
                                        in_channels,
                                        kernel_size=3,
                                        stride=2,
                                        padding=0)

    def forward(self, x):
        if self.with_conv:
            pad = (0,1,0,1)
            x = torch.nn.functional.pad(x, pad, mode="constant", value=0)
            x = self.conv(x)
        else:
            x = torch.nn.functional.avg_pool2d(x, kernel_size=2, stride=2)
        return x



class ResnetBlock(nn.Module):
    def __init__(self, *, in_channels, out_channels=None, conv_shortcut=False,
                 dropout, temb_channels=256):
        super().__init__()
        self.in_channels = in_channels
        out_channels = in_channels if out_channels is None else out_channels
        self.out_channels = out_channels
        self.use_conv_shortcut = conv_shortcut

        self.norm1 = Normalize(in_channels)
        self.conv1 = torch.nn.Conv2d(in_channels,
                                     out_channels,
                                     kernel_size=3,
                                     stride=1,
                                     padding=1)
        if temb_channels > 0:
            self.temb_proj = torch.nn.Linear(temb_channels,
                                             out_channels)
        self.norm2 = Normalize(out_channels)
        self.dropout = torch.nn.Dropout(dropout)
        self.conv2 = torch.nn.Conv2d(out_channels,
                                     out_channels,
                                     kernel_size=3,
                                     stride=1,
                                     padding=1)
        
        
        # 替换原来的conv2层
        self.spatial_conv = DepthWiseConv(out_channels,out_channels)
        self.channel_conv = SqueezeChannelConv(out_channels, out_channels)
        

        
        if self.in_channels != self.out_channels:
            if self.use_conv_shortcut:
                self.conv_shortcut = torch.nn.Conv2d(in_channels,
                                                     out_channels,
                                                     kernel_size=3,
                                                     stride=1,
                                                     padding=1)
            else:
                self.nin_shortcut = torch.nn.Conv2d(in_channels,
                                                    out_channels,
                                                    kernel_size=1,
                                                    stride=1,
                                                    padding=0)

    def forward(self, x, temb):
        h = x
        h = self.norm1(h)
        h = nonlinearity(h)
        h = self.conv1(h)

        if temb is not None:
            h = h + self.temb_proj(nonlinearity(temb))[:,:,None,None]

        h = self.norm2(h)
        h = nonlinearity(h)
        h_mid = self.dropout(h)
        # h = self.conv2(h)
        
        # 改进部分
        h_spatial = self.spatial_conv(h_mid)
        h_channel = self.channel_conv(h_mid)
        h = h_mid + h_spatial + h_channel

        if self.in_channels != self.out_channels:
            if self.use_conv_shortcut:
                x = self.conv_shortcut(x)
            else:
                x = self.nin_shortcut(x)

        return x+h


class AttnBlock(nn.Module):
    def __init__(self, in_channels):
        super().__init__()
        self.in_channels = in_channels

        self.norm = Normalize(in_channels)
        self.q = torch.nn.Conv2d(in_channels,
                                 in_channels,
                                 kernel_size=1,
                                 stride=1,
                                 padding=0)
        self.k = torch.nn.Conv2d(in_channels,
                                 in_channels,
                                 kernel_size=1,
                                 stride=1,
                                 padding=0)
        self.v = torch.nn.Conv2d(in_channels,
                                 in_channels,
                                 kernel_size=1,
                                 stride=1,
                                 padding=0)
        self.proj_out = torch.nn.Conv2d(in_channels,
                                        in_channels,
                                        kernel_size=1,
                                        stride=1,
                                        padding=0)


    def forward(self, x):
        h_ = x
        h_ = self.norm(h_)
        q = self.q(h_)
        k = self.k(h_)
        v = self.v(h_)

        # compute attention
        b,c,h,w = q.shape
        q = q.reshape(b,c,h*w)
        q = q.permute(0,2,1)   # b,hw,c
        k = k.reshape(b,c,h*w) # b,c,hw
        w_ = torch.bmm(q,k)     # b,hw,hw    w[b,i,j]=sum_c q[b,i,c]k[b,c,j]
        w_ = w_ * (int(c)**(-0.5))
        w_ = torch.nn.functional.softmax(w_, dim=2)

        # attend to values
        v = v.reshape(b,c,h*w)
        w_ = w_.permute(0,2,1)   # b,hw,hw (first hw of k, second of q)
        h_ = torch.bmm(v,w_)     # b, c,hw (hw of q) h_[b,c,j] = sum_i v[b,c,i] w_[b,i,j]
        h_ = h_.reshape(b,c,h,w)

        h_ = self.proj_out(h_)

        return x+h_


class Model(nn.Module):
    def __init__(self, *, ch, out_ch, ch_mult=(1,2,4,8), num_res_blocks,
                 attn_resolutions, dropout=0.0, resamp_with_conv=True, in_channels,
                 resolution, use_timestep=True):
        super().__init__()
        self.ch = ch
        self.temb_ch = self.ch*4
        self.num_resolutions = len(ch_mult)
        self.num_res_blocks = num_res_blocks
        self.resolution = resolution
        self.in_channels = in_channels

        self.use_timestep = use_timestep
        if self.use_timestep:
            # timestep embedding
            self.temb = nn.Module()
            self.temb.dense = nn.ModuleList([
                torch.nn.Linear(self.ch,
                                self.temb_ch),
                torch.nn.Linear(self.temb_ch,
                                self.temb_ch),
            ])

        # downsampling
        self.conv_in = torch.nn.Conv2d(in_channels,
                                       self.ch,
                                       kernel_size=3,
                                       stride=1,
                                       padding=1)

        curr_res = resolution
        in_ch_mult = (1,)+tuple(ch_mult)
        self.down = nn.ModuleList()
        for i_level in range(self.num_resolutions):
            block = nn.ModuleList()
            attn = nn.ModuleList()
            block_in = ch*in_ch_mult[i_level]
            block_out = ch*ch_mult[i_level]
            for i_block in range(self.num_res_blocks):
                block.append(ResnetBlock(in_channels=block_in,
                                         out_channels=block_out,
                                         temb_channels=self.temb_ch,
                                         dropout=dropout))
                block_in = block_out
                if curr_res in attn_resolutions:
                    attn.append(AttnBlock(block_in))
            down = nn.Module()
            down.block = block
            down.attn = attn
            if i_level != self.num_resolutions-1:
                down.downsample = Downsample(block_in, resamp_with_conv)
                curr_res = curr_res // 2
            self.down.append(down)

        # middle
        self.mid = nn.Module()
        self.mid.block_1 = ResnetBlock(in_channels=block_in,
                                       out_channels=block_in,
                                       temb_channels=self.temb_ch,
                                       dropout=dropout)
        self.mid.attn_1 = AttnBlock(block_in)
        self.mid.block_2 = ResnetBlock(in_channels=block_in,
                                       out_channels=block_in,
                                       temb_channels=self.temb_ch,
                                       dropout=dropout)

        # upsampling
        self.up = nn.ModuleList()
        for i_level in reversed(range(self.num_resolutions)):
            block = nn.ModuleList()
            attn = nn.ModuleList()
            block_out = ch*ch_mult[i_level]
            skip_in = ch*ch_mult[i_level]
            for i_block in range(self.num_res_blocks+1):
                if i_block == self.num_res_blocks:
                    skip_in = ch*in_ch_mult[i_level]
                block.append(ResnetBlock(in_channels=block_in+skip_in,
                                         out_channels=block_out,
                                         temb_channels=self.temb_ch,
                                         dropout=dropout))
                block_in = block_out
                if curr_res in attn_resolutions:
                    attn.append(AttnBlock(block_in))
            up = nn.Module()
            up.block = block
            up.attn = attn
            if i_level != 0:
                up.upsample = Upsample(block_in, resamp_with_conv)
                curr_res = curr_res * 2
            self.up.insert(0, up) # prepend to get consistent order

        # end
        self.norm_out = Normalize(block_in)
        self.conv_out = torch.nn.Conv2d(block_in,
                                        out_ch,
                                        kernel_size=3,
                                        stride=1,
                                        padding=1)


    def forward(self, x, t=None):
        #assert x.shape[2] == x.shape[3] == self.resolution

        if self.use_timestep:
            # timestep embedding
            assert t is not None
            temb = get_timestep_embedding(t, self.ch)
            temb = self.temb.dense[0](temb)
            temb = nonlinearity(temb)
            temb = self.temb.dense[1](temb)
        else:
            temb = None

        # downsampling
        hs = [self.conv_in(x)]
        for i_level in range(self.num_resolutions):
            for i_block in range(self.num_res_blocks):
                h = self.down[i_level].block[i_block](hs[-1], temb)
                if len(self.down[i_level].attn) > 0:
                    h = self.down[i_level].attn[i_block](h)
                hs.append(h)
            if i_level != self.num_resolutions-1:
                hs.append(self.down[i_level].downsample(hs[-1]))

        # middle
        h = hs[-1]
        h = self.mid.block_1(h, temb)
        h = self.mid.attn_1(h)
        h = self.mid.block_2(h, temb)

        # upsampling
        for i_level in reversed(range(self.num_resolutions)):
            for i_block in range(self.num_res_blocks+1):
                h = self.up[i_level].block[i_block](
                    torch.cat([h, hs.pop()], dim=1), temb)
                if len(self.up[i_level].attn) > 0:
                    h = self.up[i_level].attn[i_block](h)
            if i_level != 0:
                h = self.up[i_level].upsample(h)

        # end
        h = self.norm_out(h)
        h = nonlinearity(h)
        h = self.conv_out(h)
        return h


class Encoder(nn.Module):
    def __init__(self, *, ch=128, out_ch=3, ch_mult=[1,1,2,2,4], num_res_blocks=2,
                 attn_resolutions=[16], dropout=0.0, resamp_with_conv=True, in_channels=3,
                 resolution=256, z_channels=256, double_z=False, **ignore_kwargs):
        super().__init__()
        self.ch = ch
        self.temb_ch = 256
        self.num_resolutions = len(ch_mult)
        self.num_res_blocks = num_res_blocks
        self.resolution = resolution
        self.in_channels = in_channels

        # downsampling
        self.conv_in = torch.nn.Conv2d(in_channels,
                                       self.ch,
                                       kernel_size=3,
                                       stride=1,
                                       padding=1)

        curr_res = resolution
        in_ch_mult = (1,)+tuple(ch_mult)
        self.down = nn.ModuleList()
        for i_level in range(self.num_resolutions):
            block = nn.ModuleList()
            attn = nn.ModuleList()
            block_in = ch*in_ch_mult[i_level]
            block_out = ch*ch_mult[i_level]
            for i_block in range(self.num_res_blocks):
                block.append(ResnetBlock(in_channels=block_in,
                                         out_channels=block_out,
                                         temb_channels=self.temb_ch,
                                         dropout=dropout))
                block_in = block_out
                if curr_res in attn_resolutions:
                    attn.append(AttnBlock(block_in))
            down = nn.Module()
            down.block = block
            down.attn = attn
            if i_level != self.num_resolutions-1:
                down.downsample = Downsample(block_in, resamp_with_conv)
                curr_res = curr_res // 2
            self.down.append(down)

        # middle
        self.mid = nn.Module()
        self.mid.block_1 = ResnetBlock(in_channels=block_in,
                                       out_channels=block_in,
                                       temb_channels=self.temb_ch,
                                       dropout=dropout)
        self.mid.attn_1 = AttnBlock(block_in)
        self.mid.block_2 = ResnetBlock(in_channels=block_in,
                                       out_channels=block_in,
                                       temb_channels=self.temb_ch,
                                       dropout=dropout)

        # end
        self.norm_out = Normalize(block_in)
        self.conv_out = torch.nn.Conv2d(block_in,
                                        2*z_channels if double_z else z_channels,
                                        kernel_size=3,
                                        stride=1,
                                        padding=1)
                                        
        # weak label的emb变量
        self.weak_emb = WeakImgEmb(3, 3)


    def forward(self, x, weak_img):
        #assert x.shape[2] == x.shape[3] == self.resolution, "{}, {}, {}".format(x.shape[2], x.shape[3], self.resolution)

        # timestep embedding
        temb = self.weak_emb(weak_img) # B, 16*16*3-->B,256
        # temb = None

        # downsampling
        hs = [self.conv_in(x)]
        for i_level in range(self.num_resolutions):
            for i_block in range(self.num_res_blocks):
                h = self.down[i_level].block[i_block](hs[-1], temb)
                if len(self.down[i_level].attn) > 0:
                    h = self.down[i_level].attn[i_block](h)
                hs.append(h)
            if i_level != self.num_resolutions-1:
                hs.append(self.down[i_level].downsample(hs[-1]))

        # middle
        h = hs[-1]
        h = self.mid.block_1(h, temb)
        h = self.mid.attn_1(h)
        h = self.mid.block_2(h, temb)

        # end
        h = self.norm_out(h)
        h = nonlinearity(h)
        h = self.conv_out(h)
        return h


class Decoder(nn.Module):
    def __init__(self, *, ch=128, out_ch=3, ch_mult=[1,1,2,2,4], num_res_blocks=2,
                 attn_resolutions=[16], dropout=0.0, resamp_with_conv=True, in_channels=3,
                 resolution=256, z_channels=256, double_z=False, give_pre_end=False, **ignorekwargs):
        super().__init__()
        self.ch = ch
        self.temb_ch = 0
        self.num_resolutions = len(ch_mult)
        self.num_res_blocks = num_res_blocks
        self.resolution = resolution
        self.in_channels = in_channels
        self.give_pre_end = give_pre_end

        # compute in_ch_mult, block_in and curr_res at lowest res
        in_ch_mult = (1,)+tuple(ch_mult)
        block_in = ch*ch_mult[self.num_resolutions-1]
        curr_res = resolution // 2**(self.num_resolutions-1)
        self.z_shape = (1,z_channels,curr_res,curr_res)
        print("Working with z of shape {} = {} dimensions.".format(
            self.z_shape, np.prod(self.z_shape)))

        # z to block_in
        self.conv_in = torch.nn.Conv2d(z_channels,
                                       block_in,
                                       kernel_size=3,
                                       stride=1,
                                       padding=1)

        # middle
        self.mid = nn.Module()
        self.mid.block_1 = ResnetBlock(in_channels=block_in,
                                       out_channels=block_in,
                                       temb_channels=self.temb_ch,
                                       dropout=dropout)
        self.mid.attn_1 = AttnBlock(block_in)
        self.mid.block_2 = ResnetBlock(in_channels=block_in,
                                       out_channels=block_in,
                                       temb_channels=self.temb_ch,
                                       dropout=dropout)

        # upsampling
        self.up = nn.ModuleList()
        for i_level in reversed(range(self.num_resolutions)):
            block = nn.ModuleList()
            attn = nn.ModuleList()
            block_out = ch*ch_mult[i_level]
            for i_block in range(self.num_res_blocks+1):
                block.append(ResnetBlock(in_channels=block_in,
                                         out_channels=block_out,
                                         temb_channels=self.temb_ch,
                                         dropout=dropout))
                block_in = block_out
                if curr_res in attn_resolutions:
                    attn.append(AttnBlock(block_in))
            up = nn.Module()
            up.block = block
            up.attn = attn
            if i_level != 0:
                up.upsample = Upsample(block_in, resamp_with_conv)
                curr_res = curr_res * 2
            self.up.insert(0, up) # prepend to get consistent order

        # end
        self.norm_out = Normalize(block_in)
        self.conv_out = torch.nn.Conv2d(block_in,
                                        out_ch,
                                        kernel_size=3,
                                        stride=1,
                                        padding=1)

    def forward(self, z):
        #assert z.shape[1:] == self.z_shape[1:]
        self.last_z_shape = z.shape

        # timestep embedding
        temb = None

        # z to block_in
        h = self.conv_in(z)

        # middle
        h = self.mid.block_1(h, temb)
        h = self.mid.attn_1(h)
        h = self.mid.block_2(h, temb)

        # upsampling
        for i_level in reversed(range(self.num_resolutions)):
            for i_block in range(self.num_res_blocks+1):
                h = self.up[i_level].block[i_block](h, temb)
                if len(self.up[i_level].attn) > 0:
                    h = self.up[i_level].attn[i_block](h)
            if i_level != 0:
                h = self.up[i_level].upsample(h)

        # end
        if self.give_pre_end:
            return h

        h = self.norm_out(h)
        h = nonlinearity(h)
        h = self.conv_out(h)
        return h


class VUNet(nn.Module):
    def __init__(self, *, ch, out_ch, ch_mult=(1,2,4,8), num_res_blocks,
                 attn_resolutions, dropout=0.0, resamp_with_conv=True,
                 in_channels, c_channels,
                 resolution, z_channels, use_timestep=False, **ignore_kwargs):
        super().__init__()
        self.ch = ch
        self.temb_ch = self.ch*4
        self.num_resolutions = len(ch_mult)
        self.num_res_blocks = num_res_blocks
        self.resolution = resolution

        self.use_timestep = use_timestep
        if self.use_timestep:
            # timestep embedding
            self.temb = nn.Module()
            self.temb.dense = nn.ModuleList([
                torch.nn.Linear(self.ch,
                                self.temb_ch),
                torch.nn.Linear(self.temb_ch,
                                self.temb_ch),
            ])

        # downsampling
        self.conv_in = torch.nn.Conv2d(c_channels,
                                       self.ch,
                                       kernel_size=3,
                                       stride=1,
                                       padding=1)

        curr_res = resolution
        in_ch_mult = (1,)+tuple(ch_mult)
        self.down = nn.ModuleList()
        for i_level in range(self.num_resolutions):
            block = nn.ModuleList()
            attn = nn.ModuleList()
            block_in = ch*in_ch_mult[i_level]
            block_out = ch*ch_mult[i_level]
            for i_block in range(self.num_res_blocks):
                block.append(ResnetBlock(in_channels=block_in,
                                         out_channels=block_out,
                                         temb_channels=self.temb_ch,
                                         dropout=dropout))
                block_in = block_out
                if curr_res in attn_resolutions:
                    attn.append(AttnBlock(block_in))
            down = nn.Module()
            down.block = block
            down.attn = attn
            if i_level != self.num_resolutions-1:
                down.downsample = Downsample(block_in, resamp_with_conv)
                curr_res = curr_res // 2
            self.down.append(down)

        self.z_in = torch.nn.Conv2d(z_channels,
                                    block_in,
                                    kernel_size=1,
                                    stride=1,
                                    padding=0)
        # middle
        self.mid = nn.Module()
        self.mid.block_1 = ResnetBlock(in_channels=2*block_in,
                                       out_channels=block_in,
                                       temb_channels=self.temb_ch,
                                       dropout=dropout)
        self.mid.attn_1 = AttnBlock(block_in)
        self.mid.block_2 = ResnetBlock(in_channels=block_in,
                                       out_channels=block_in,
                                       temb_channels=self.temb_ch,
                                       dropout=dropout)

        # upsampling
        self.up = nn.ModuleList()
        for i_level in reversed(range(self.num_resolutions)):
            block = nn.ModuleList()
            attn = nn.ModuleList()
            block_out = ch*ch_mult[i_level]
            skip_in = ch*ch_mult[i_level]
            for i_block in range(self.num_res_blocks+1):
                if i_block == self.num_res_blocks:
                    skip_in = ch*in_ch_mult[i_level]
                block.append(ResnetBlock(in_channels=block_in+skip_in,
                                         out_channels=block_out,
                                         temb_channels=self.temb_ch,
                                         dropout=dropout))
                block_in = block_out
                if curr_res in attn_resolutions:
                    attn.append(AttnBlock(block_in))
            up = nn.Module()
            up.block = block
            up.attn = attn
            if i_level != 0:
                up.upsample = Upsample(block_in, resamp_with_conv)
                curr_res = curr_res * 2
            self.up.insert(0, up) # prepend to get consistent order

        # end
        self.norm_out = Normalize(block_in)
        self.conv_out = torch.nn.Conv2d(block_in,
                                        out_ch,
                                        kernel_size=3,
                                        stride=1,
                                        padding=1)


    def forward(self, x, z):
        #assert x.shape[2] == x.shape[3] == self.resolution

        if self.use_timestep:
            # timestep embedding
            assert t is not None
            temb = get_timestep_embedding(t, self.ch)
            temb = self.temb.dense[0](temb)
            temb = nonlinearity(temb)
            temb = self.temb.dense[1](temb)
        else:
            temb = None

        # downsampling
        hs = [self.conv_in(x)]
        for i_level in range(self.num_resolutions):
            for i_block in range(self.num_res_blocks):
                h = self.down[i_level].block[i_block](hs[-1], temb)
                if len(self.down[i_level].attn) > 0:
                    h = self.down[i_level].attn[i_block](h)
                hs.append(h)
            if i_level != self.num_resolutions-1:
                hs.append(self.down[i_level].downsample(hs[-1]))

        # middle
        h = hs[-1]
        z = self.z_in(z)
        h = torch.cat((h,z),dim=1)
        h = self.mid.block_1(h, temb)
        h = self.mid.attn_1(h)
        h = self.mid.block_2(h, temb)

        # upsampling
        for i_level in reversed(range(self.num_resolutions)):
            for i_block in range(self.num_res_blocks+1):
                h = self.up[i_level].block[i_block](
                    torch.cat([h, hs.pop()], dim=1), temb)
                if len(self.up[i_level].attn) > 0:
                    h = self.up[i_level].attn[i_block](h)
            if i_level != 0:
                h = self.up[i_level].upsample(h)

        # end
        h = self.norm_out(h)
        h = nonlinearity(h)
        h = self.conv_out(h)
        return h


class SimpleDecoder(nn.Module):
    def __init__(self, in_channels, out_channels, *args, **kwargs):
        super().__init__()
        self.model = nn.ModuleList([nn.Conv2d(in_channels, in_channels, 1),
                                     ResnetBlock(in_channels=in_channels,
                                                 out_channels=2 * in_channels,
                                                 temb_channels=0, dropout=0.0),
                                     ResnetBlock(in_channels=2 * in_channels,
                                                out_channels=4 * in_channels,
                                                temb_channels=0, dropout=0.0),
                                     ResnetBlock(in_channels=4 * in_channels,
                                                out_channels=2 * in_channels,
                                                temb_channels=0, dropout=0.0),
                                     nn.Conv2d(2*in_channels, in_channels, 1),
                                     Upsample(in_channels, with_conv=True)])
        # end
        self.norm_out = Normalize(in_channels)
        self.conv_out = torch.nn.Conv2d(in_channels,
                                        out_channels,
                                        kernel_size=3,
                                        stride=1,
                                        padding=1)

    def forward(self, x):
        for i, layer in enumerate(self.model):
            if i in [1,2,3]:
                x = layer(x, None)
            else:
                x = layer(x)

        h = self.norm_out(x)
        h = nonlinearity(h)
        x = self.conv_out(h)
        return x


class UpsampleDecoder(nn.Module):
    def __init__(self, in_channels, out_channels, ch, num_res_blocks, resolution,
                 ch_mult=(2,2), dropout=0.0):
        super().__init__()
        # upsampling
        self.temb_ch = 0
        self.num_resolutions = len(ch_mult)
        self.num_res_blocks = num_res_blocks
        block_in = in_channels
        curr_res = resolution // 2 ** (self.num_resolutions - 1)
        self.res_blocks = nn.ModuleList()
        self.upsample_blocks = nn.ModuleList()
        for i_level in range(self.num_resolutions):
            res_block = []
            block_out = ch * ch_mult[i_level]
            for i_block in range(self.num_res_blocks + 1):
                res_block.append(ResnetBlock(in_channels=block_in,
                                         out_channels=block_out,
                                         temb_channels=self.temb_ch,
                                         dropout=dropout))
                block_in = block_out
            self.res_blocks.append(nn.ModuleList(res_block))
            if i_level != self.num_resolutions - 1:
                self.upsample_blocks.append(Upsample(block_in, True))
                curr_res = curr_res * 2

        # end
        self.norm_out = Normalize(block_in)
        self.conv_out = torch.nn.Conv2d(block_in,
                                        out_channels,
                                        kernel_size=3,
                                        stride=1,
                                        padding=1)

    def forward(self, x):
        # upsampling
        h = x
        for k, i_level in enumerate(range(self.num_resolutions)):
            for i_block in range(self.num_res_blocks + 1):
                h = self.res_blocks[i_level][i_block](h, None)
            if i_level != self.num_resolutions - 1:
                h = self.upsample_blocks[k](h)
        h = self.norm_out(h)
        h = nonlinearity(h)
        h = self.conv_out(h)
        return h