model_hw.py

from copy import deepcopy
import pdb
import torch
import torch.nn as nn
import torch.utils.checkpoint as checkpoint
from timm.models.layers import DropPath, to_2tuple, trunc_normal_
import torch.nn.functional as F
from einops import rearrange, repeat
from einops.layers.torch import Rearrange
import math
import numpy as np
import time
from torch import einsum
from torchvision.utils import save_image

from mimo_modules.MIMOUNet import EBlock, DBlock


class PredictorLG(nn.Module):
    """ Image to Patch Embedding
    """

    def __init__(self, embed_dim=384):
        super().__init__()
        self.in_conv = nn.Sequential(
            nn.Linear(embed_dim, embed_dim),
            nn.GELU(),
            nn.LayerNorm(embed_dim),
            nn.Linear(embed_dim, embed_dim),
        )

        self.out_conv = nn.Sequential(
            nn.Linear(embed_dim, embed_dim // 2),
            nn.GELU(),
            nn.Linear(embed_dim // 2, embed_dim // 4),
            nn.GELU(),
            nn.Linear(embed_dim // 4, 2),
        )

    def forward(self, x):
        x = self.in_conv(x)
        B, N, C = x.size()
        local_x = x[:, :, :C // 2]
        global_x = (x[:, :, C // 2:]).mean(dim=1, keepdim=True)
        x = torch.cat([local_x, global_x.expand(B, N, C // 2)], dim=-1)
        return self.out_conv(x)


class FastLeFF(nn.Module):

    def __init__(self, dim=32, hidden_dim=128, act_layer=nn.GELU, drop=0.):
        super().__init__()

        from torch_dwconv import depthwise_conv2d, DepthwiseConv2d

        self.linear1 = nn.Sequential(nn.Linear(dim, hidden_dim),
                                     act_layer())
        self.dwconv = nn.Sequential(DepthwiseConv2d(hidden_dim, hidden_dim, kernel_size=3, stride=1, padding=1),
                                    act_layer())
        self.linear2 = nn.Sequential(nn.Linear(hidden_dim, dim))
        self.dim = dim
        self.hidden_dim = hidden_dim

    def forward(self, x):
        # bs x h x w x c
        x = self.linear1(x).permute(0, 3, 1, 2)

        # spatial restore
        x = self.dwconv(x).permute(0, 2, 3, 1)

        # flaten
        x = self.linear2(x)

        return x

    def flops(self, H, W):
        flops = 0
        # fc1
        flops += H * W * self.dim * self.hidden_dim
        # dwconv
        flops += H * W * self.hidden_dim * 3 * 3
        # fc2
        flops += H * W * self.hidden_dim * self.dim
        print("LeFF:{%.2f}" % (flops / 1e9))
        return flops


def conv(in_channels, out_channels, kernel_size, bias=False, stride=1):
    return nn.Conv2d(
        in_channels, out_channels, kernel_size,
        padding=(kernel_size // 2), bias=bias, stride=stride)


## Supervised Attention Module
class SAM(nn.Module):
    def __init__(self, n_feat, kernel_size=3, bias=True):
        super(SAM, self).__init__()
        self.conv1 = conv(n_feat, n_feat, kernel_size, bias=bias)
        self.conv2 = conv(n_feat, 3, kernel_size, bias=bias)
        self.conv3 = conv(3, n_feat, kernel_size, bias=bias)

    def forward(self, x, x_img):
        x1 = self.conv1(x)
        img = self.conv2(x) + x_img
        x2 = torch.sigmoid(self.conv3(img))
        x1 = x1 * x2
        x1 = x1 + x
        return x1, img


#########################################
class ConvBlock(nn.Module):
    def __init__(self, in_channel, out_channel, strides=1):
        super(ConvBlock, self).__init__()
        self.strides = strides
        self.in_channel = in_channel
        self.out_channel = out_channel
        self.block = nn.Sequential(
            nn.Conv2d(in_channel, out_channel, kernel_size=3, stride=strides, padding=1),
            nn.LeakyReLU(inplace=True),
            nn.Conv2d(out_channel, out_channel, kernel_size=3, stride=strides, padding=1),
            nn.LeakyReLU(inplace=True),
        )
        self.conv11 = nn.Conv2d(in_channel, out_channel, kernel_size=1, stride=strides, padding=0)

    def forward(self, x):
        out1 = self.block(x)
        out2 = self.conv11(x)
        out = out1 + out2
        return out

    def flops(self, H, W):
        flops = H * W * self.in_channel * self.out_channel * (
                3 * 3 + 1) + H * W * self.out_channel * self.out_channel * 3 * 3
        return flops


class eca_layer_1d(nn.Module):
    """Constructs a ECA module.
    Args:
        channel: Number of channels of the input feature map
        k_size: Adaptive selection of kernel size
    """

    def __init__(self, channel, k_size=3):
        super(eca_layer_1d, self).__init__()
        self.avg_pool = nn.AdaptiveAvgPool1d(1)
        self.conv = nn.Conv1d(1, 1, kernel_size=k_size, padding=(k_size - 1) // 2, bias=False)
        self.sigmoid = nn.Sigmoid()
        self.channel = channel
        self.k_size = k_size

    def forward(self, x):
        # b hw c
        # feature descriptor on the global spatial information
        y = self.avg_pool(x.transpose(-1, -2))

        # Two different branches of ECA module
        y = self.conv(y.transpose(-1, -2))

        # Multi-scale information fusion
        y = self.sigmoid(y)

        return x * y.expand_as(x)

    def flops(self):
        flops = 0
        flops += self.channel * self.channel * self.k_size

        return flops


class SepConv2d(torch.nn.Module):
    def __init__(self,
                 in_channels,
                 out_channels,
                 kernel_size,
                 stride=1,
                 padding=0,
                 dilation=1, act_layer=nn.ReLU):
        super(SepConv2d, self).__init__()
        self.depthwise = torch.nn.Conv2d(in_channels,
                                         in_channels,
                                         kernel_size=kernel_size,
                                         stride=stride,
                                         padding=padding,
                                         dilation=dilation,
                                         groups=in_channels)
        self.pointwise = torch.nn.Conv2d(in_channels, out_channels, kernel_size=1)
        self.act_layer = act_layer() if act_layer is not None else nn.Identity()
        self.in_channels = in_channels
        self.out_channels = out_channels
        self.kernel_size = kernel_size
        self.stride = stride

    def forward(self, x):
        x = self.depthwise(x)
        x = self.act_layer(x)
        x = self.pointwise(x)
        return x

    def flops(self, HW):
        flops = 0
        flops += HW * self.in_channels * self.kernel_size ** 2 / self.stride ** 2
        flops += HW * self.in_channels * self.out_channels
        print("SeqConv2d:{%.2f}" % (flops / 1e9))
        return flops


######## Embedding for q,k,v ########
class ConvProjection(nn.Module):
    def __init__(self, dim, heads=8, dim_head=64, kernel_size=3, q_stride=1, k_stride=1, v_stride=1, dropout=0.,
                 last_stage=False, bias=True):
        super().__init__()

        inner_dim = dim_head * heads
        self.heads = heads
        pad = (kernel_size - q_stride) // 2
        self.to_q = SepConv2d(dim, inner_dim, kernel_size, q_stride, pad, bias)
        self.to_k = SepConv2d(dim, inner_dim, kernel_size, k_stride, pad, bias)
        self.to_v = SepConv2d(dim, inner_dim, kernel_size, v_stride, pad, bias)

    def forward(self, x, attn_kv=None):
        b, n, c, h = *x.shape, self.heads
        l = int(math.sqrt(n))
        w = int(math.sqrt(n))

        attn_kv = x if attn_kv is None else attn_kv
        x = rearrange(x, 'b (l w) c -> b c l w', l=l, w=w)
        attn_kv = rearrange(attn_kv, 'b (l w) c -> b c l w', l=l, w=w)
        # print(attn_kv)
        q = self.to_q(x)
        q = rearrange(q, 'b (h d) l w -> b h (l w) d', h=h)

        k = self.to_k(attn_kv)
        v = self.to_v(attn_kv)
        k = rearrange(k, 'b (h d) l w -> b h (l w) d', h=h)
        v = rearrange(v, 'b (h d) l w -> b h (l w) d', h=h)
        return q, k, v

    def flops(self, q_L, kv_L=None):
        kv_L = kv_L or q_L
        flops = 0
        flops += self.to_q.flops(q_L)
        flops += self.to_k.flops(kv_L)
        flops += self.to_v.flops(kv_L)
        return flops


class LinearProjection(nn.Module):
    def __init__(self, dim, heads=8, dim_head=64, dropout=0., bias=True):
        super().__init__()
        inner_dim = dim_head * heads
        self.heads = heads
        self.to_q = nn.Linear(dim, inner_dim, bias=bias)
        self.to_kv = nn.Linear(dim, inner_dim * 2, bias=bias)
        self.dim = dim
        self.inner_dim = inner_dim

    def forward(self, x, attn_kv=None):
        B_, N, C = x.shape
        if attn_kv is not None:
            attn_kv = attn_kv.unsqueeze(0).repeat(B_, 1, 1)
        else:
            attn_kv = x
        N_kv = attn_kv.size(1)
        q = self.to_q(x).reshape(B_, N, 1, self.heads, C // self.heads).permute(2, 0, 3, 1, 4)
        kv = self.to_kv(attn_kv).reshape(B_, N_kv, 2, self.heads, C // self.heads).permute(2, 0, 3, 1, 4)
        q = q[0]
        k, v = kv[0], kv[1]
        return q, k, v

    def flops(self, q_L, kv_L=None):
        kv_L = kv_L or q_L
        flops = q_L * self.dim * self.inner_dim + kv_L * self.dim * self.inner_dim * 2
        return flops


#########################################
########### window-based self-attention #############
class WindowAttention(nn.Module):
    def __init__(self, dim, win_size, num_heads, token_projection='linear', qkv_bias=True, qk_scale=None, attn_drop=0.,
                 proj_drop=0.):

        super().__init__()
        self.dim = dim
        self.win_size = win_size  # Wh, Ww
        self.num_heads = num_heads
        head_dim = dim // num_heads
        self.scale = qk_scale or head_dim ** -0.5

        # define a parameter table of relative position bias
        self.relative_position_bias_table = nn.Parameter(
            torch.zeros((2 * win_size[0] - 1) * (2 * win_size[1] - 1), num_heads))  # 2*Wh-1 * 2*Ww-1, nH

        # get pair-wise relative position index for each token inside the window
        coords_h = torch.arange(self.win_size[0])  # [0,...,Wh-1]
        coords_w = torch.arange(self.win_size[1])  # [0,...,Ww-1]
        coords = torch.stack(torch.meshgrid([coords_h, coords_w]))  # 2, Wh, Ww
        coords_flatten = torch.flatten(coords, 1)  # 2, Wh*Ww
        relative_coords = coords_flatten[:, :, None] - coords_flatten[:, None, :]  # 2, Wh*Ww, Wh*Ww
        relative_coords = relative_coords.permute(1, 2, 0).contiguous()  # Wh*Ww, Wh*Ww, 2
        relative_coords[:, :, 0] += self.win_size[0] - 1  # shift to start from 0
        relative_coords[:, :, 1] += self.win_size[1] - 1
        relative_coords[:, :, 0] *= 2 * self.win_size[1] - 1
        relative_position_index = relative_coords.sum(-1)  # Wh*Ww, Wh*Ww
        self.register_buffer("relative_position_index", relative_position_index)
        trunc_normal_(self.relative_position_bias_table, std=.02)

        if token_projection == 'conv':
            self.qkv = ConvProjection(dim, num_heads, dim // num_heads, bias=qkv_bias)
        elif token_projection == 'linear':
            self.qkv = LinearProjection(dim, num_heads, dim // num_heads, bias=qkv_bias)
        else:
            raise Exception("Projection error!")

        self.token_projection = token_projection
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(dim, dim)
        self.proj_drop = nn.Dropout(proj_drop)

        self.softmax = nn.Softmax(dim=-1)

    def forward(self, x, attn_kv=None, mask=None):
        B_, N, C = x.shape
        q, k, v = self.qkv(x, attn_kv)
        q = q * self.scale
        attn = (q @ k.transpose(-2, -1))

        relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
            self.win_size[0] * self.win_size[1], self.win_size[0] * self.win_size[1], -1)  # Wh*Ww,Wh*Ww,nH
        relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww
        ratio = attn.size(-1) // relative_position_bias.size(-1)
        relative_position_bias = repeat(relative_position_bias, 'nH l c -> nH l (c d)', d=ratio)

        attn = attn + relative_position_bias.unsqueeze(0)

        if mask is not None:
            nW = mask.shape[0]
            mask = repeat(mask, 'nW m n -> nW m (n d)', d=ratio)
            attn = attn.view(B_ // nW, nW, self.num_heads, N, N * ratio) + mask.unsqueeze(1).unsqueeze(0)
            attn = attn.view(-1, self.num_heads, N, N * ratio)
            attn = self.softmax(attn)
        else:
            attn = self.softmax(attn)

        attn = self.attn_drop(attn)

        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
        x = self.proj(x)
        x = self.proj_drop(x)
        return x

    def extra_repr(self) -> str:
        return f'dim={self.dim}, win_size={self.win_size}, num_heads={self.num_heads}'

    def flops(self, H, W):
        # calculate flops for 1 window with token length of N
        # print(N, self.dim)
        flops = 0
        N = self.win_size[0] * self.win_size[1]
        nW = H * W / N
        # qkv = self.qkv(x)
        # flops += N * self.dim * 3 * self.dim
        flops += self.qkv.flops(H * W, H * W)

        # attn = (q @ k.transpose(-2, -1))

        flops += nW * self.num_heads * N * (self.dim // self.num_heads) * N
        #  x = (attn @ v)
        flops += nW * self.num_heads * N * N * (self.dim // self.num_heads)

        # x = self.proj(x)
        flops += nW * N * self.dim * self.dim
        print("W-MSA:{%.2f}" % (flops / 1e9))
        return flops


########### self-attention #############
class Attention(nn.Module):
    def __init__(self, dim, num_heads, token_projection='linear', qkv_bias=True, qk_scale=None, attn_drop=0.,
                 proj_drop=0.):

        super().__init__()
        self.dim = dim
        self.num_heads = num_heads
        head_dim = dim // num_heads
        self.scale = qk_scale or head_dim ** -0.5

        self.qkv = LinearProjection(dim, num_heads, dim // num_heads, bias=qkv_bias)

        self.token_projection = token_projection
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(dim, dim)
        self.proj_drop = nn.Dropout(proj_drop)

        self.softmax = nn.Softmax(dim=-1)

    def forward(self, x, attn_kv=None, mask=None):
        B_, N, C = x.shape
        q, k, v = self.qkv(x, attn_kv)
        q = q * self.scale
        attn = (q @ k.transpose(-2, -1))

        # relative_position_bias = self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
        #     self.win_size[0] * self.win_size[1], self.win_size[0] * self.win_size[1], -1)  # Wh*Ww,Wh*Ww,nH
        # relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous()  # nH, Wh*Ww, Wh*Ww
        # ratio = attn.size(-1)//relative_position_bias.size(-1)
        # relative_position_bias = repeat(relative_position_bias, 'nH l c -> nH l (c d)', d = ratio)

        # attn = attn + relative_position_bias.unsqueeze(0)

        if mask is not None:
            nW = mask.shape[0]
            # mask = repeat(mask, 'nW m n -> nW m (n d)',d = ratio)
            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(1).unsqueeze(0)
            attn = attn.view(-1, self.num_heads, N, N)
            attn = self.softmax(attn)
        else:
            attn = self.softmax(attn)

        attn = self.attn_drop(attn)

        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
        x = self.proj(x)
        x = self.proj_drop(x)
        return x

    def extra_repr(self) -> str:
        return f'dim={self.dim}, num_heads={self.num_heads}'

    def flops(self, q_num, kv_num):
        # calculate flops for 1 window with token length of N
        # print(N, self.dim)
        flops = 0
        # N = self.win_size[0]*self.win_size[1]
        # nW = H*W/N
        # qkv = self.qkv(x)
        # flops += N * self.dim * 3 * self.dim
        flops += self.qkv.flops(q_num, kv_num)
        # attn = (q @ k.transpose(-2, -1))

        flops += self.num_heads * q_num * (self.dim // self.num_heads) * kv_num
        #  x = (attn @ v)
        flops += self.num_heads * q_num * (self.dim // self.num_heads) * kv_num

        # x = self.proj(x)
        flops += q_num * self.dim * self.dim
        print("MCA:{%.2f}" % (flops / 1e9))
        return flops


#########################################
########### feed-forward network #############
class Mlp(nn.Module):
    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.):
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features
        self.fc1 = nn.Linear(in_features, hidden_features)
        self.act = act_layer()
        self.fc2 = nn.Linear(hidden_features, out_features)
        self.drop = nn.Dropout(drop)
        self.in_features = in_features
        self.hidden_features = hidden_features
        self.out_features = out_features

    def forward(self, x):
        x = self.fc1(x)
        x = self.act(x)
        x = self.drop(x)
        x = self.fc2(x)
        x = self.drop(x)
        return x

    def flops(self, H, W):
        flops = 0
        # fc1
        flops += H * W * self.in_features * self.hidden_features
        # fc2
        flops += H * W * self.hidden_features * self.out_features
        print("MLP:{%.2f}" % (flops / 1e9))
        return flops


class LeFF(nn.Module):
    def __init__(self, dim=32, hidden_dim=128, act_layer=nn.GELU, drop=0., use_eca=False):
        super().__init__()
        self.linear1 = nn.Sequential(nn.Linear(dim, hidden_dim),
                                     act_layer())
        self.dwconv = nn.Sequential(
            nn.Conv2d(hidden_dim, hidden_dim, groups=hidden_dim, kernel_size=3, stride=1, padding=1),
            act_layer())
        self.linear2 = nn.Sequential(nn.Linear(hidden_dim, dim))
        self.dim = dim
        self.hidden_dim = hidden_dim
        self.eca = eca_layer_1d(dim) if use_eca else nn.Identity()

    def forward(self, x):
        # bs x h x w x c
        x = self.linear1(x).permute(0, 3, 1, 2)

        x = self.dwconv(x).permute(0, 2, 3, 1)

        x = self.linear2(x)
        x = self.eca(x)

        return x

    def flops(self, H, W):
        flops = 0
        # fc1
        flops += H * W * self.dim * self.hidden_dim
        # dwconv
        flops += H * W * self.hidden_dim * 3 * 3
        # fc2
        flops += H * W * self.hidden_dim * self.dim
        print("LeFF:{%.2f}" % (flops / 1e9))
        # eca 
        if hasattr(self.eca, 'flops'):
            flops += self.eca.flops()
        return flops


class LeFFLocal(LeFF):
    def __init__(self, dim=32, hidden_dim=128, act_layer=nn.GELU, drop=0., use_eca=False):
        super().__init__()
        self.linear1 = nn.Sequential(nn.Linear(dim, hidden_dim),
                                     act_layer())
        self.dwconv = nn.Sequential(
            nn.Conv2d(hidden_dim, hidden_dim, groups=hidden_dim, kernel_size=3, stride=1, padding=1,
                      padding_mode='reflect'),
            act_layer())
        self.linear2 = nn.Sequential(nn.Linear(hidden_dim, dim))
        self.dim = dim
        self.hidden_dim = hidden_dim
        self.eca = eca_layer_1d(dim) if use_eca else nn.Identity()


#########################################
########### window operation#############
def window_partition(x, win_size, dilation_rate=1):
    B, H, W, C = x.shape
    if dilation_rate != 1:
        x = x.permute(0, 3, 1, 2)  # B, C, H, W
        assert type(dilation_rate) is int, 'dilation_rate should be a int'
        x = F.unfold(x, kernel_size=win_size, dilation=dilation_rate, padding=4 * (dilation_rate - 1),
                     stride=win_size)  # B, C*Wh*Ww, H/Wh*W/Ww
        windows = x.permute(0, 2, 1).contiguous().view(-1, C, win_size, win_size)  # B' ,C ,Wh ,Ww
        windows = windows.permute(0, 2, 3, 1).contiguous()  # B' ,Wh ,Ww ,C
    else:
        x = x.view(B, H // win_size, win_size, W // win_size, win_size, C)
        windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, win_size, win_size, C)  # B' ,Wh ,Ww ,C
    return windows


def window_reverse(windows, win_size, H, W, dilation_rate=1):
    # B' ,Wh ,Ww ,C
    B = int(windows.shape[0] / (H * W / win_size / win_size))
    x = windows.view(B, H // win_size, W // win_size, win_size, win_size, -1)
    if dilation_rate != 1:
        x = windows.permute(0, 5, 3, 4, 1, 2).contiguous()  # B, C*Wh*Ww, H/Wh*W/Ww
        x = F.fold(x, (H, W), kernel_size=win_size, dilation=dilation_rate, padding=4 * (dilation_rate - 1),
                   stride=win_size)
    else:
        x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1)
    return x


#########################################
# Downsample Block
class Downsample(nn.Module):
    def __init__(self, in_channel, out_channel):
        super(Downsample, self).__init__()
        self.conv = nn.Sequential(
            nn.Conv2d(in_channel, out_channel, kernel_size=4, stride=2, padding=1),
        )
        self.in_channel = in_channel
        self.out_channel = out_channel

    def forward(self, x):
        x = x.permute(0, 3, 1, 2)
        out = self.conv(x).permute(0, 2, 3, 1)  # B H*W C
        return out

    def flops(self, H, W):
        flops = 0
        # conv
        flops += H / 2 * W / 2 * self.in_channel * self.out_channel * 4 * 4
        print("Downsample:{%.2f}" % (flops / 1e9))
        return flops


# Upsample Block
class Upsample(nn.Module):
    def __init__(self, in_channel, out_channel):
        super(Upsample, self).__init__()
        self.deconv = nn.Sequential(
            nn.ConvTranspose2d(in_channel, out_channel, kernel_size=2, stride=2),
        )
        self.in_channel = in_channel
        self.out_channel = out_channel

    def forward(self, x):
        x = x.permute(0, 3, 1, 2)
        out = self.deconv(x).permute(0, 2, 3, 1).contiguous()  # B H*W C
        return out

    def flops(self, H, W):
        flops = 0
        # conv
        flops += H * 2 * W * 2 * self.in_channel * self.out_channel * 2 * 2
        print("Upsample:{%.2f}" % (flops / 1e9))
        return flops


# Input Projection
class InputProj(nn.Module):
    def __init__(self, in_channel=3, out_channel=64, kernel_size=3, stride=1, norm_layer=None, act_layer=nn.LeakyReLU):
        super().__init__()
        self.proj = nn.Sequential(
            nn.Conv2d(in_channel, out_channel, kernel_size=3, stride=stride, padding=kernel_size // 2),
            act_layer(inplace=True)
        )
        if norm_layer is not None:
            self.norm = norm_layer(out_channel)
        else:
            self.norm = None
        self.in_channel = in_channel
        self.out_channel = out_channel

    def forward(self, x):
        B, C, H, W = x.shape
        x = self.proj(x).permute(0, 2, 3, 1).contiguous()  # B H*W C
        if self.norm is not None:
            x = self.norm(x)
        return x

    def flops(self, H, W):
        flops = 0
        # conv
        flops += H * W * self.in_channel * self.out_channel * 3 * 3

        if self.norm is not None:
            flops += H * W * self.out_channel
        print("Input_proj:{%.2f}" % (flops / 1e9))
        return flops


# Output Projection
class OutputProj(nn.Module):
    def __init__(self, in_channel=64, out_channel=3, kernel_size=3, stride=1, norm_layer=None, act_layer=None):
        super().__init__()
        self.proj = nn.Sequential(
            nn.Conv2d(in_channel, out_channel, kernel_size=3, stride=stride, padding=kernel_size // 2),
        )
        if act_layer is not None:
            self.proj.add_module(act_layer(inplace=True))
        if norm_layer is not None:
            self.norm = norm_layer(out_channel)
        else:
            self.norm = None
        self.in_channel = in_channel
        self.out_channel = out_channel

    def forward(self, x):
        x = x.permute(0, 3, 1, 2)
        x = self.proj(x)
        if self.norm is not None:
            x = self.norm(x)
        return x

    def flops(self, H, W):
        flops = 0
        # conv
        flops += H * W * self.in_channel * self.out_channel * 3 * 3

        if self.norm is not None:
            flops += H * W * self.out_channel
        print("Output_proj:{%.2f}" % (flops / 1e9))
        return flops


#########################################
########### LeWinTransformer #############
class LeWinTransformerBlock(nn.Module):
    def __init__(self, dim, input_resolution, num_heads, win_size=8, shift_size=0,
                 mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0., drop_path=0.,
                 act_layer=nn.GELU, norm_layer=nn.LayerNorm, token_projection='linear', token_mlp='leff',
                 modulator=False, cross_modulator=False, prune=False, predict_prune_score=False):
        super().__init__()
        self.dim = dim
        self.input_resolution = input_resolution
        self.num_heads = num_heads
        self.win_size = win_size
        self.shift_size = shift_size
        self.mlp_ratio = mlp_ratio
        self.token_mlp = token_mlp
        self.predict_prune_score = predict_prune_score
        self.prune = prune
        if min(self.input_resolution) <= self.win_size:
            self.shift_size = 0
            self.win_size = min(self.input_resolution)
        assert 0 <= self.shift_size < self.win_size, "shift_size must in 0-win_size"

        if modulator:
            self.modulator = nn.Embedding(win_size * win_size, dim)  # modulator
        else:
            self.modulator = None

        if cross_modulator:
            self.cross_modulator = nn.Embedding(win_size * win_size, dim)  # cross_modulator
            self.cross_attn = Attention(dim, num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop,
                                        proj_drop=drop,
                                        token_projection=token_projection, )
            self.norm_cross = norm_layer(dim)
        else:
            self.cross_modulator = None

        self.norm1 = norm_layer(dim)
        self.attn = WindowAttention(
            dim, win_size=to_2tuple(self.win_size), num_heads=num_heads,
            qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop,
            token_projection=token_projection)

        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
        self.norm2 = norm_layer(dim)
        mlp_hidden_dim = int(dim * mlp_ratio)
        if token_mlp in ['ffn', 'mlp']:
            self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)
        elif token_mlp == 'leff':
            self.mlp = LeFF(dim, mlp_hidden_dim, act_layer=act_layer, drop=drop)
        elif token_mlp == 'lefflocal':
            self.mlp = LeFFLocal(dim, mlp_hidden_dim, act_layer=act_layer, drop=drop)
        elif token_mlp == 'fastleff':
            self.mlp = FastLeFF(dim, mlp_hidden_dim, act_layer=act_layer, drop=drop)
        else:
            raise Exception("FFN error!")

        if prune and predict_prune_score:
            self.score_predictor = PredictorLG(self.dim)

    def with_pos_embed(self, tensor, pos):
        return tensor if pos is None else tensor + pos

    def extra_repr(self) -> str:
        return f"dim={self.dim}, input_resolution={self.input_resolution}, num_heads={self.num_heads}, " \
               f"win_size={self.win_size}, shift_size={self.shift_size}, mlp_ratio={self.mlp_ratio},modulator={self.modulator}"

    def forward(self, x, mask=None, policy=None):
        B, H, W, C = x.shape
        x = x.reshape(B, H * W, C)

        ## input mask
        if mask != None:
            input_mask = F.interpolate(mask, size=(H, W)).permute(0, 2, 3, 1)
            input_mask_windows = window_partition(input_mask, self.win_size)  # nW, win_size, win_size, 1
            attn_mask = input_mask_windows.view(-1, self.win_size * self.win_size)  # nW, win_size*win_size
            attn_mask = attn_mask.unsqueeze(2) * attn_mask.unsqueeze(1)  # nW, win_size*win_size, win_size*win_size
            attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(attn_mask == 0, float(0.0))
        else:
            attn_mask = None

        ## shift mask
        if self.shift_size > 0:
            # calculate attention mask for SW-MSA
            shift_mask = torch.zeros((1, H, W, 1)).type_as(x)
            h_slices = (slice(0, -self.win_size),
                        slice(-self.win_size, -self.shift_size),
                        slice(-self.shift_size, None))
            w_slices = (slice(0, -self.win_size),
                        slice(-self.win_size, -self.shift_size),
                        slice(-self.shift_size, None))
            cnt = 0
            for h in h_slices:
                for w in w_slices:
                    shift_mask[:, h, w, :] = cnt
                    cnt += 1
            shift_mask_windows = window_partition(shift_mask, self.win_size)  # nW, win_size, win_size, 1
            shift_mask_windows = shift_mask_windows.view(-1, self.win_size * self.win_size)  # nW, win_size*win_size
            shift_attn_mask = shift_mask_windows.unsqueeze(1) - shift_mask_windows.unsqueeze(
                2)  # nW, win_size*win_size, win_size*win_size
            shift_attn_mask = shift_attn_mask.masked_fill(shift_attn_mask != 0, float(-100.0)).masked_fill(
                shift_attn_mask == 0, float(0.0))
            attn_mask = attn_mask + shift_attn_mask if attn_mask is not None else shift_attn_mask

        if self.cross_modulator is not None:
            shortcut = x
            x_cross = self.norm_cross(x)
            x_cross = self.cross_attn(x, self.cross_modulator.weight)
            x = shortcut + x_cross

        if self.prune:
            x_ret, policy_cur, pred_score = self.forward_prune(x, H, W, policy, attn_mask)
            return x_ret, policy_cur, pred_score
            # if not self.training and self.prune:
            #     print(111, x_ret[0, :3, :3, :3])
            # return self.forward_prune(x, H, W, policy, attn_mask)

        shortcut = x
        x = self.norm1(x)
        x = x.view(B, H, W, C)

        # cyclic shift
        if self.shift_size > 0:
            shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
        else:
            shifted_x = x

        # partition windows
        x_windows = window_partition(shifted_x, self.win_size)  # nW*B, win_size, win_size, C  N*C->C
        x_windows = x_windows.view(-1, self.win_size * self.win_size, C)  # nW*B, win_size*win_size, C

        # with_modulator
        if self.modulator is not None:
            wmsa_in = self.with_pos_embed(x_windows, self.modulator.weight)
        else:
            wmsa_in = x_windows

        # W-MSA/SW-MSA
        attn_windows = self.attn(wmsa_in, mask=attn_mask)  # nW*B, win_size*win_size, C

        # merge windows
        attn_windows = attn_windows.view(-1, self.win_size, self.win_size, C)

        if self.token_mlp == 'lefflocal':
            # cyclic shift
            shortcut = shortcut.view(B, H, W, C)
            if self.shift_size > 0:
                shortcut = torch.roll(shortcut, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
            shortcut = window_partition(shortcut, self.win_size)  # B H' W' C
            # FFN
            x = shortcut + self.drop_path(attn_windows)
            x = x + self.drop_path(self.mlp(self.norm2(x)))
            x = window_reverse(x, self.win_size, H, W)  # B H' W' C
            # reverse cyclic shift
            if self.shift_size > 0:
                x = torch.roll(x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
            x = x.view(B, H, W, C)
        else:
            shifted_x = window_reverse(attn_windows, self.win_size, H, W)  # B H' W' C

            # reverse cyclic shift
            if self.shift_size > 0:
                x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
            else:
                x = shifted_x
            x = x.view(B, H * W, C)

            # FFN
            x = shortcut + self.drop_path(x)
            x = x.view(B, H, W, C)
            x = x + self.drop_path(self.mlp(self.norm2(x)))
        del attn_mask
        if self.prune:
            return x, policy_cur, pred_score
        return x

    def forward_prune(self, x, H, W, policy_prev, attn_mask):
        shortcut = x
        B, _, C = x.shape
        # training_mode = True
        training_mode = self.training
        # select tokens
        if self.predict_prune_score:
            pred_score_ = pred_score = self.score_predictor(x).reshape(B, H, W, 2)
            if self.shift_size > 0:
                pred_score_ = torch.roll(pred_score, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
            pred_score_win = window_partition(pred_score_, self.win_size).mean([1, 2])  # B*nW, win_size, win_size, 2
            # pred_score_win = torch.randn_like(pred_score_win)  # For testing: force not to prune
            # pred_score_win[:, 0] = 1e9  # For testing: force not to prune
            if training_mode:
                hard_keep_decision = F.gumbel_softmax(pred_score_win, hard=True)[:, 0]
                policy_cur = hard_keep_decision * policy_prev
            else:
                hard_keep_decision = F.softmax(pred_score_win, -1)[..., 0] > 0.5 #0.5
                policy_cur = hard_keep_decision * policy_prev
        else:
            pred_score = None
            policy_cur = policy_prev

        if policy_cur.sum() == 0:
            return shortcut.view(B, H, W, C), policy_cur, pred_score

        x = self.norm1(x)
        x = x.view(B, H, W, C)

        # cyclic shift
        if self.shift_size > 0:
            shifted_x = torch.roll(x, shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
            shifted_shortcut = torch.roll(
                shortcut.view(B, H, W, C), shifts=(-self.shift_size, -self.shift_size), dims=(1, 2))
        else:
            shifted_x = x
            shifted_shortcut = shortcut.view(B, H, W, C)

        # partition windows
        x_windows = window_partition(shifted_x, self.win_size)  # nW*B, win_size, win_size, C  N*C->C
        if training_mode:
            x_windows = x_windows * policy_cur[:, None, None, None]
        else:
            x_windows = x_windows[policy_cur > 0]
            if self.shift_size > 0:
                attn_mask = attn_mask[policy_cur > 0]
        x_windows = x_windows.view(-1, self.win_size * self.win_size, C)  # nW*B, win_size*win_size, C
        # with_modulator
        if self.modulator is not None:
            wmsa_in = self.with_pos_embed(x_windows, self.modulator.weight)
        else:
            wmsa_in = x_windows

        # W-MSA/SW-MSA
        attn_windows = self.attn(wmsa_in, mask=attn_mask)  # nW*B, win_size*win_size, C
        attn_windows = attn_windows.view(-1, self.win_size, self.win_size, C)

        if self.token_mlp == 'lefflocal':
            shortcut = window_partition(shifted_shortcut, self.win_size)
            if training_mode:
                policy_windows_ = policy_cur[:, None, None, None]
                x = shortcut + self.drop_path(attn_windows) * policy_windows_
                x = x + self.drop_path(self.mlp(self.norm2(x))) * policy_windows_
                x = window_reverse(x, self.win_size, H, W)
            else:
                x = attn_windows + shortcut[policy_cur > 0]  # [nS, w, w, C]
                x = x + self.mlp(self.norm2(x))
                x_ = shortcut.clone()
                x_[policy_cur > 0] = x
                x = window_reverse(x_, self.win_size, H, W)
            if self.shift_size > 0:
                x = torch.roll(x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
        else:
            if training_mode:
                shifted_x = window_reverse(attn_windows, self.win_size, H, W)
                # reverse cyclic shift
                if self.shift_size > 0:
                    x = torch.roll(shifted_x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
                else:
                    x = shifted_x

                policy_windows_ = policy_cur[:, None, None, None].repeat(1, self.win_size, self.win_size, 1)
                policy_windows_ = window_reverse(policy_windows_, self.win_size, H, W)
                # x = x.view(B, H * W, C)
                # policy_windows_ = policy_windows_.view(B, H * W, 1)
                # x = shortcut + self.drop_path(x) * policy_windows_
                x = shifted_shortcut + self.drop_path(x) * policy_windows_
                x = x + self.drop_path(self.mlp(self.norm2(x))) * policy_windows_
            else:
                # shortcut = window_partition(shifted_shortcut, self.win_size)
                # x = attn_windows + shortcut[policy_cur > 0]  # [nS, w, w, C]
                # x = x + self.mlp(self.norm2(x))
                # x_ = shortcut.clone()
                # x_[policy_cur > 0] = x
                # x = window_reverse(x_, self.win_size, H, W)
                # if self.shift_size > 0:
                #     x = torch.roll(x, shifts=(self.shift_size, self.shift_size), dims=(1, 2))
                shortcut = window_partition(shifted_shortcut, self.win_size)
                x = attn_windows + shortcut[policy_cur > 0]  # [nS, w, w, C]
                x_ = shortcut.clone()
                x_[policy_cur > 0] = x
                x = window_reverse(x_, self.win_size, H, W)
                # print(x.shape)
                # pdb.set_trace()
                x = x + self.mlp(self.norm2(x))
                # reverse cyclic shift
                if self.shift_size > 0:
                    x = torch.roll(x, shifts=(self.shift_size, self.shift_size), dims=(1, 2)) 
                
        x = x.view(B, H, W, C)
        # print(">>> 1", pred_score_win[:3], x[0, :3, :3, :5], policy_cur, x.shape)
        # if self.shift_size > 0:
        #     exit()
        del attn_mask
        return x, policy_cur, pred_score

    def flops(self):
        flops = 0
        H, W = self.input_resolution

        if self.cross_modulator is not None:
            flops += self.dim * H * W
            flops += self.cross_attn.flops(H * W, self.win_size * self.win_size)

        # norm1
        flops += self.dim * H * W
        # W-MSA/SW-MSA
        flops += self.attn.flops(H, W)
        # norm2
        flops += self.dim * H * W
        # mlp
        flops += self.mlp.flops(H, W)
        # print("LeWin:{%.2f}"%(flops/1e9))
        return flops


#########################################
########### Basic layer ################
class AdaWPT(nn.Module):
    def __init__(self, dim, output_dim, input_resolution, depth, num_heads, win_size,
                 mlp_ratio=4., qkv_bias=True, qk_scale=None, drop=0., attn_drop=0.,
                 drop_path=0., norm_layer=nn.LayerNorm, use_checkpoint=False,
                 token_projection='linear', token_mlp='ffn', shift_flag=True,
                 modulator=False, cross_modulator=False, prune=False):

        super().__init__()
        self.dim = dim
        self.input_resolution = input_resolution
        self.depth = depth
        self.prune = prune
        self.use_checkpoint = use_checkpoint
        self.win_size = win_size
        self.shift_flag = shift_flag
        # build blocks
        if shift_flag:
            self.blocks = nn.ModuleList([
                LeWinTransformerBlock(dim=dim, input_resolution=input_resolution,
                                      num_heads=num_heads, win_size=win_size,
                                      shift_size=0 if (i % 2 == 0) else win_size // 2,
                                      mlp_ratio=mlp_ratio,
                                      qkv_bias=qkv_bias, qk_scale=qk_scale,
                                      drop=drop, attn_drop=attn_drop,
                                      drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,
                                      norm_layer=norm_layer, token_projection=token_projection, token_mlp=token_mlp,
                                      modulator=modulator, cross_modulator=cross_modulator,
                                      prune=prune, predict_prune_score=i <= 1)
                for i in range(depth)])
        else:
            self.blocks = nn.ModuleList([
                LeWinTransformerBlock(dim=dim, input_resolution=input_resolution,
                                      num_heads=num_heads, win_size=win_size,
                                      shift_size=0,
                                      mlp_ratio=mlp_ratio,
                                      qkv_bias=qkv_bias, qk_scale=qk_scale,
                                      drop=drop, attn_drop=attn_drop,
                                      drop_path=drop_path[i] if isinstance(drop_path, list) else drop_path,
                                      norm_layer=norm_layer, token_projection=token_projection, token_mlp=token_mlp,
                                      modulator=modulator, cross_modulator=cross_modulator,
                                      prune=prune, predict_prune_score=i == 0)
                for i in range(depth)])

    def extra_repr(self) -> str:
        return f"dim={self.dim}, input_resolution={self.input_resolution}, depth={self.depth}"

    def forward(self, x, mask=None):
        if self.prune:
            return self.forward_prune(x, mask)
        for blk in self.blocks:
            if self.use_checkpoint:
                x = checkpoint.checkpoint(blk, x)
            else:
                x = blk(x, mask)
        return x

    def forward_prune(self, x, mask=None):
        B, H, W, _ = x.shape
        if self.shift_flag:
            policy_windows = [
                torch.ones([B * H // self.win_size * W // self.win_size], dtype=x.dtype, device=x.device),
                torch.ones([B * H // self.win_size * W // self.win_size], dtype=x.dtype, device=x.device),
            ]
        else:
            policy_windows = [
                torch.ones([B * H // self.win_size * W // self.win_size], dtype=x.dtype, device=x.device),
            ]
        pred_score_list = []
        decision_list = []
        for i, blk in enumerate(self.blocks):
            if self.use_checkpoint:
                x = checkpoint.checkpoint(blk, x)
            else:
                x, policy_windows[i % len(policy_windows)], score = blk(
                    x, mask, policy_windows[i % len(policy_windows)])
                if score is not None:
                    pred_score_list.append(score)
                    decision_list.append(
                        policy_windows[i % len(policy_windows)].reshape(
                            B, 1, H // self.win_size, W // self.win_size).detach())
        return x, pred_score_list, decision_list

    def flops(self):
        flops = 0
        for blk in self.blocks:
            flops += blk.flops()
        return flops






class LMD(nn.Module):
    def __init__(self, img_size=256, in_chans=3, dd_in=3,
                 embed_dim=32, depths=[2, 2, 2, 2, 2, 2, 2, 2, 2], num_heads=[1, 2, 4, 8, 16, 16, 8, 4, 2],
                 win_size=8, mlp_ratio=4., qkv_bias=True, qk_scale=None,
                 drop_rate=0., attn_drop_rate=0., drop_path_rate=0.1,
                 norm_layer=nn.LayerNorm, patch_norm=True,
                 use_checkpoint=False, token_projection='linear', token_mlp='leff',
                 dowsample=Downsample, upsample=Upsample, shift_flag=True, modulator=False,
                 cross_modulator=False, prune_loc=[0, 0, 1, 1, 1, 1, 1, 0, 0], **kwargs):
        super().__init__()

        self.num_enc_layers = len(depths) // 2
        self.num_dec_layers = len(depths) // 2
        self.embed_dim = embed_dim
        self.patch_norm = patch_norm
        self.mlp_ratio = mlp_ratio
        self.token_projection = token_projection
        self.mlp = token_mlp
        self.win_size = win_size
        self.reso = img_size
        self.pos_drop = nn.Dropout(p=drop_rate)
        self.dd_in = dd_in

        # stochastic depth
        enc_dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths[:self.num_enc_layers]))]
        conv_dpr = [drop_path_rate] * depths[4]
        dec_dpr = enc_dpr[::-1]
        self.dec_dpr = dec_dpr

        # build layers

        # Input/Output
        self.input_proj = InputProj(in_channel=dd_in, out_channel=embed_dim, kernel_size=3, stride=1,
                                    act_layer=nn.LeakyReLU)
        self.output_proj = OutputProj(in_channel=2 * embed_dim, out_channel=in_chans + 1, kernel_size=3, stride=1)

        # Encoder
        self.encoderlayer_0 = AdaWPT(dim=embed_dim,
                                                output_dim=embed_dim,
                                                input_resolution=(img_size,
                                                                  img_size),
                                                depth=depths[0],
                                                num_heads=num_heads[0],
                                                win_size=win_size,
                                                mlp_ratio=self.mlp_ratio,
                                                qkv_bias=qkv_bias, qk_scale=qk_scale,
                                                drop=drop_rate, attn_drop=attn_drop_rate,
                                                drop_path=enc_dpr[sum(depths[:0]):sum(depths[:1])],
                                                norm_layer=norm_layer,
                                                use_checkpoint=use_checkpoint,
                                                token_projection=token_projection, token_mlp=token_mlp,
                                                shift_flag=shift_flag,
                                                prune=prune_loc[0] == 1)
        self.dowsample_0 = dowsample(embed_dim, embed_dim * 2)
        self.encoderlayer_1 = AdaWPT(dim=embed_dim * 2,
                                                output_dim=embed_dim * 2,
                                                input_resolution=(img_size // 2,
                                                                  img_size // 2),
                                                depth=depths[1],
                                                num_heads=num_heads[1],
                                                win_size=win_size,
                                                mlp_ratio=self.mlp_ratio,
                                                qkv_bias=qkv_bias, qk_scale=qk_scale,
                                                drop=drop_rate, attn_drop=attn_drop_rate,
                                                drop_path=enc_dpr[sum(depths[:1]):sum(depths[:2])],
                                                norm_layer=norm_layer,
                                                use_checkpoint=use_checkpoint,
                                                token_projection=token_projection, token_mlp=token_mlp,
                                                shift_flag=shift_flag,
                                                prune=prune_loc[1] == 1)
        self.dowsample_1 = dowsample(embed_dim * 2, embed_dim * 4)
        self.encoderlayer_2 = AdaWPT(dim=embed_dim * 4,
                                                output_dim=embed_dim * 4,
                                                input_resolution=(img_size // (2 ** 2),
                                                                  img_size // (2 ** 2)),
                                                depth=depths[2],
                                                num_heads=num_heads[2],
                                                win_size=win_size,
                                                mlp_ratio=self.mlp_ratio,
                                                qkv_bias=qkv_bias, qk_scale=qk_scale,
                                                drop=drop_rate, attn_drop=attn_drop_rate,
                                                drop_path=enc_dpr[sum(depths[:2]):sum(depths[:3])],
                                                norm_layer=norm_layer,
                                                use_checkpoint=use_checkpoint,
                                                token_projection=token_projection, token_mlp=token_mlp,
                                                shift_flag=shift_flag,
                                                prune=prune_loc[2] == 1)
        self.dowsample_2 = dowsample(embed_dim * 4, embed_dim * 8)
        self.encoderlayer_3 = AdaWPT(dim=embed_dim * 8,
                                                output_dim=embed_dim * 8,
                                                input_resolution=(img_size // (2 ** 3),
                                                                  img_size // (2 ** 3)),
                                                depth=depths[3],
                                                num_heads=num_heads[3],
                                                win_size=win_size,
                                                mlp_ratio=self.mlp_ratio,
                                                qkv_bias=qkv_bias, qk_scale=qk_scale,
                                                drop=drop_rate, attn_drop=attn_drop_rate,
                                                drop_path=enc_dpr[sum(depths[:3]):sum(depths[:4])],
                                                norm_layer=norm_layer,
                                                use_checkpoint=use_checkpoint,
                                                token_projection=token_projection, token_mlp=token_mlp,
                                                shift_flag=shift_flag,
                                                prune=prune_loc[3] == 1)
        self.dowsample_3 = dowsample(embed_dim * 8, embed_dim * 16)

        # Bottleneck
        self.conv = AdaWPT(dim=embed_dim * 16,
                                      output_dim=embed_dim * 16,
                                      input_resolution=(img_size // (2 ** 4),
                                                        img_size // (2 ** 4)),
                                      depth=depths[4],
                                      num_heads=num_heads[4],
                                      win_size=win_size,
                                      mlp_ratio=self.mlp_ratio,
                                      qkv_bias=qkv_bias, qk_scale=qk_scale,
                                      drop=drop_rate, attn_drop=attn_drop_rate,
                                      drop_path=conv_dpr,
                                      norm_layer=norm_layer,
                                      use_checkpoint=use_checkpoint,
                                      token_projection=token_projection, token_mlp=token_mlp, shift_flag=shift_flag,
                                      prune=prune_loc[4] == 1)

        # Decoder
        self.upsample_0 = upsample(embed_dim * 16, embed_dim * 8)
        self.decoderlayer_0 = AdaWPT(dim=embed_dim * 16,
                                                output_dim=embed_dim * 16,
                                                input_resolution=(img_size // (2 ** 3),
                                                                  img_size // (2 ** 3)),
                                                depth=depths[5],
                                                num_heads=num_heads[5],
                                                win_size=win_size,
                                                mlp_ratio=self.mlp_ratio,
                                                qkv_bias=qkv_bias, qk_scale=qk_scale,
                                                drop=drop_rate, attn_drop=attn_drop_rate,
                                                drop_path=dec_dpr[:depths[5]],
                                                norm_layer=norm_layer,
                                                use_checkpoint=use_checkpoint,
                                                token_projection=token_projection, token_mlp=token_mlp,
                                                shift_flag=shift_flag,
                                                modulator=modulator, cross_modulator=cross_modulator,
                                                prune=prune_loc[5] == 1)
        self.upsample_1 = upsample(embed_dim * 16, embed_dim * 4)
        self.decoderlayer_1 = AdaWPT(dim=embed_dim * 8,
                                                output_dim=embed_dim * 8,
                                                input_resolution=(img_size // (2 ** 2),
                                                                  img_size // (2 ** 2)),
                                                depth=depths[6],
                                                num_heads=num_heads[6],
                                                win_size=win_size,
                                                mlp_ratio=self.mlp_ratio,
                                                qkv_bias=qkv_bias, qk_scale=qk_scale,
                                                drop=drop_rate, attn_drop=attn_drop_rate,
                                                drop_path=dec_dpr[sum(depths[5:6]):sum(depths[5:7])],
                                                norm_layer=norm_layer,
                                                use_checkpoint=use_checkpoint,
                                                token_projection=token_projection, token_mlp=token_mlp,
                                                shift_flag=shift_flag,
                                                modulator=modulator, cross_modulator=cross_modulator,
                                                prune=prune_loc[6] == 1)
        self.upsample_2 = upsample(embed_dim * 8, embed_dim * 2)
        self.decoderlayer_2 = AdaWPT(dim=embed_dim * 4,
                                                output_dim=embed_dim * 4,
                                                input_resolution=(img_size // 2,
                                                                  img_size // 2),
                                                depth=depths[7],
                                                num_heads=num_heads[7],
                                                win_size=win_size,
                                                mlp_ratio=self.mlp_ratio,
                                                qkv_bias=qkv_bias, qk_scale=qk_scale,
                                                drop=drop_rate, attn_drop=attn_drop_rate,
                                                drop_path=dec_dpr[sum(depths[5:7]):sum(depths[5:8])],
                                                norm_layer=norm_layer,
                                                use_checkpoint=use_checkpoint,
                                                token_projection=token_projection, token_mlp=token_mlp,
                                                shift_flag=shift_flag,
                                                modulator=modulator, cross_modulator=cross_modulator,
                                                prune=prune_loc[7] == 1)
        self.upsample_3 = upsample(embed_dim * 4, embed_dim)
        self.decoderlayer_3 = AdaWPT(dim=embed_dim * 2,
                                                output_dim=embed_dim * 2,
                                                input_resolution=(img_size,
                                                                  img_size),
                                                depth=depths[8],
                                                num_heads=num_heads[8],
                                                win_size=win_size,
                                                mlp_ratio=self.mlp_ratio,
                                                qkv_bias=qkv_bias, qk_scale=qk_scale,
                                                drop=drop_rate, attn_drop=attn_drop_rate,
                                                drop_path=dec_dpr[sum(depths[5:8]):sum(depths[5:9])],
                                                norm_layer=norm_layer,
                                                use_checkpoint=use_checkpoint,
                                                token_projection=token_projection, token_mlp=token_mlp,
                                                shift_flag=shift_flag,
                                                modulator=modulator, cross_modulator=cross_modulator,
                                                prune=prune_loc[8] == 1)
        self.prune_loc = prune_loc
        self.apply(self._init_weights)

    def _init_weights(self, m):
        if isinstance(m, nn.Linear):
            trunc_normal_(m.weight, std=.02)
            if isinstance(m, nn.Linear) and m.bias is not None:
                nn.init.constant_(m.bias, 0)
        elif isinstance(m, nn.LayerNorm):
            nn.init.constant_(m.bias, 0)
            nn.init.constant_(m.weight, 1.0)

    @torch.jit.ignore
    def no_weight_decay(self):
        return {'absolute_pos_embed'}

    @torch.jit.ignore
    def no_weight_decay_keywords(self):
        return {'relative_position_bias_table'}

    def extra_repr(self) -> str:
        return f"embed_dim={self.embed_dim}, token_projection={self.token_projection}, token_mlp={self.mlp},win_size={self.win_size}"

    def forward(self, x, mask=None):
        # Input Projection
        y = self.input_proj(x)
        y = self.pos_drop(y)
        pred_score_lists = []
        decision_lists = []
        # Encoder
        conv0 = self.encoderlayer_0(y, mask=mask)
        if self.prune_loc[0]:
            conv0, pred_score_list, decision_list = conv0
            pred_score_lists += pred_score_list
            decision_lists += decision_list
        pool0 = self.dowsample_0(conv0)
        
        conv1 = self.encoderlayer_1(pool0, mask=mask)
        if self.prune_loc[1]:
            conv1, pred_score_list, decision_list = conv1
            pred_score_lists += pred_score_list
            decision_lists += decision_list
        pool1 = self.dowsample_1(conv1)
        
        conv2 = self.encoderlayer_2(pool1, mask=mask)
        if self.prune_loc[2]:
            conv2, pred_score_list, decision_list = conv2
            pred_score_lists += pred_score_list
            decision_lists += decision_list
        pool2 = self.dowsample_2(conv2)
        
        conv3 = self.encoderlayer_3(pool2, mask=mask)
        if self.prune_loc[3]:
            conv3, pred_score_list, decision_list = conv3
            pred_score_lists += pred_score_list
            decision_lists += decision_list
        pool3 = self.dowsample_3(conv3)
        
        # Bottleneck
        conv4 = self.conv(pool3, mask=mask)
        if self.prune_loc[4]:
            conv4, pred_score_list, decision_list = conv4
            pred_score_lists += pred_score_list
            decision_lists += decision_list
       
        # Decoder
        up0 = self.upsample_0(conv4)
        deconv0 = torch.cat([up0, conv3], -1)
        deconv0 = self.decoderlayer_0(deconv0, mask=mask)
        if self.prune_loc[5]:
            deconv0, pred_score_list, decision_list = deconv0
            pred_score_lists += pred_score_list
            decision_lists += decision_list
        
        up1 = self.upsample_1(deconv0)
        deconv1 = torch.cat([up1, conv2], -1)
        deconv1 = self.decoderlayer_1(deconv1, mask=mask)
        if self.prune_loc[6]:
            deconv1, pred_score_list, decision_list = deconv1
            pred_score_lists += pred_score_list
            decision_lists += decision_list
       
        up2 = self.upsample_2(deconv1)
        deconv2 = torch.cat([up2, conv1], -1)
        deconv2 = self.decoderlayer_2(deconv2, mask=mask)
        if self.prune_loc[7]:
            deconv2, pred_score_list, decision_list = deconv2
            pred_score_lists += pred_score_list
            decision_lists += decision_list
        
        up3 = self.upsample_3(deconv2)
        deconv3 = torch.cat([up3, conv0], -1)
        deconv3 = self.decoderlayer_3(deconv3, mask=mask)
        if self.prune_loc[8]:
            deconv3, pred_score_list, decision_list = deconv3
            pred_score_lists += pred_score_list
            decision_lists += decision_list
        
        # Output Projection
        y = self.output_proj(deconv3)
        gate_x = y[:, -1:]
        y = y[:, :3]

        gate_x = F.sigmoid(gate_x[:, :1])
        gate_x = gate_x.clamp_min(0.1)

        # latents = [pool0, pool1,pool2, pool3, conv4, deconv0, deconv1, deconv2, deconv3]
        return x + y * gate_x if self.dd_in == 3 else y, gate_x, pred_score_lists, decision_lists