lvvit_l2_3keep.py

import torch
import torch.nn as nn
import numpy as np
from functools import partial
import torch.nn.init as init
import torch.nn.functional as F
import math
from timm.models.layers import DropPath, to_2tuple

import torch
import torch.nn as nn

from timm.models.layers import trunc_normal_
import numpy as np
import json

from utils import batch_index_select

file = 'lvvit_l2_score.json'

def _cfg(url='', **kwargs):
    return {
        'url': url,
        'num_classes': 1000, 'input_size': (3, 224, 224), 'pool_size': None,
        'crop_pct': .9, 'interpolation': 'bicubic',
        'mean': (0.485, 0.456, 0.406), 'std': (0.229, 0.224, 0.225),
        'classifier': 'head',
        **kwargs
    }

default_cfgs = {
    'LV_ViT_Tiny': _cfg(),
    'LV_ViT': _cfg(),
    'LV_ViT_Medium': _cfg(crop_pct=1.0),
    'LV_ViT_Large': _cfg(crop_pct=1.0),
}

DROPOUT_FLOPS = 4
LAYER_NORM_FLOPS = 5
ACTIVATION_FLOPS = 8
SOFTMAX_FLOPS = 5

class GroupLinear(nn.Module):
    '''
    Group Linear operator 
    '''
    def __init__(self, in_planes, out_channels,groups=1, bias=True):
        super(GroupLinear, self).__init__()
        assert in_planes%groups==0
        assert out_channels%groups==0
        self.in_dim = in_planes
        self.out_dim = out_channels
        self.groups=groups
        self.bias = bias
        self.group_in_dim = int(self.in_dim/self.groups)
        self.group_out_dim = int(self.out_dim/self.groups)

        self.group_weight = nn.Parameter(torch.zeros(self.groups, self.group_in_dim, self.group_out_dim))
        self.group_bias=nn.Parameter(torch.zeros(self.out_dim))

    def forward(self, x):
        t,b,d=x.size()
        x = x.view(t,b,self.groups,int(d/self.groups))
        out = torch.einsum('tbgd,gdf->tbgf', (x, self.group_weight)).reshape(t,b,self.out_dim)+self.group_bias
        return out
    def extra_repr(self):
        s = ('{in_dim}, {out_dim}')
        if self.groups != 1:
            s += ', groups={groups}'
        if self.bias is None:
            s += ', bias=False'
        return s.format(**self.__dict__)


class Mlp(nn.Module):
    '''
    MLP with support to use group linear operator
    '''
    def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0., group=1):
        super().__init__()
        out_features = out_features or in_features
        hidden_features = hidden_features or in_features
        if group==1:
            self.fc1 = nn.Linear(in_features, hidden_features)
            self.fc2 = nn.Linear(hidden_features, out_features)
        else:
            self.fc1 = GroupLinear(in_features, hidden_features,group)
            self.fc2 = GroupLinear(hidden_features, out_features,group)
        self.act = act_layer()

        self.drop = nn.Dropout(drop)

    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

class GroupNorm(nn.Module):
    def __init__(self, num_groups, embed_dim, eps=1e-5, affine=True):
        super().__init__()
        self.gn = nn.GroupNorm(num_groups, embed_dim,eps,affine)

    def forward(self, x):
        B,T,C = x.shape
        x = x.view(B*T,C)
        x = self.gn(x)
        x = x.view(B,T,C)
        return x


class Attention(nn.Module):
    '''
    Multi-head self-attention
    from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
    with some modification to support different num_heads and head_dim.
    '''
    def __init__(self, dim, num_heads=8, head_dim=None, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0.):
        super().__init__()
        self.num_heads = num_heads
        if head_dim is not None:
            self.head_dim=head_dim
        else:
            head_dim = dim // num_heads
            self.head_dim = head_dim
        self.scale = qk_scale or head_dim ** -0.5

        self.qkv = nn.Linear(dim, self.head_dim* self.num_heads * 3, bias=qkv_bias)
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(self.head_dim* self.num_heads, dim)
        self.proj_drop = nn.Dropout(proj_drop)

    def softmax_with_policy(self, attn, policy, eps=1e-6):
        B, N, _ = policy.size()
        B, H, N, N = attn.size()
        attn_policy = policy.reshape(B, 1, 1, N)  # * policy.reshape(B, 1, N, 1)
        eye = torch.eye(N, dtype=attn_policy.dtype, device=attn_policy.device).view(1, 1, N, N)
        attn_policy = attn_policy + (1.0 - attn_policy) * eye
        max_att = torch.max(attn, dim=-1, keepdim=True)[0]
        attn = attn - max_att
        # attn = attn.exp_() * attn_policy
        # return attn / attn.sum(dim=-1, keepdim=True)

        # for stable training
        attn = attn.to(torch.float32).exp_() * attn_policy.to(torch.float32)
        attn = (attn + eps/N) / (attn.sum(dim=-1, keepdim=True) + eps)
        return attn.type_as(max_att)

    def forward(self, x, policy, padding_mask=None):
        B, N, C = x.shape
        qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, self.head_dim).permute(2, 0, 3, 1, 4)
        # B,heads,N,C/heads 
        q, k, v = qkv[0], qkv[1], qkv[2]
        
        # trick here to make q@k.t more stable
        attn = ((q * self.scale) @ k.transpose(-2, -1))
        if padding_mask is not None:
            # attn = attn.view(B, self.num_heads, N, N)
            # attn = attn.masked_fill(
            #     padding_mask.unsqueeze(1).unsqueeze(2).to(torch.bool),
            #     float("-inf"),
            # )
            # attn_float = attn.softmax(dim=-1, dtype=torch.float32)
            # attn = attn_float.type_as(attn)
            raise NotImplementedError
        else:
            if policy is None:
                attn = attn.softmax(dim=-1)
            elif not self.training:
                attn = self.softmax_with_policy(attn, policy, 0)
            else:
                attn = self.softmax_with_policy(attn, policy, 1e-6)
        attn = self.attn_drop(attn)

        x = (attn @ v).transpose(1, 2).reshape(B, N, self.head_dim* self.num_heads)
        x = self.proj(x)
        x = self.proj_drop(x)
        return x
        
class Block(nn.Module):
    '''
    Pre-layernorm transformer block
    '''
    def __init__(self, dim, num_heads, head_dim=None, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0.,
                 drop_path=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm, group=1, skip_lam=1.):
        super().__init__()
        self.dim = dim
        self.mlp_hidden_dim = int(dim * mlp_ratio)
        self.skip_lam = skip_lam

        self.norm1 = norm_layer(dim)
        self.attn = Attention(
            dim, num_heads=num_heads, head_dim=head_dim, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
        self.norm2 = norm_layer(dim)
        self.mlp = Mlp(in_features=dim, hidden_features=self.mlp_hidden_dim, act_layer=act_layer, drop=drop, group=group)

    def forward(self, x, policy=None, padding_mask=None):
        x = x + self.drop_path(self.attn(self.norm1(x), policy, padding_mask))/self.skip_lam
        x = x + self.drop_path(self.mlp(self.norm2(x)))/self.skip_lam
        return x

    def flops(self, s):
        heads = self.attn.num_heads
        h = self.dim
        i = self.mlp_hidden_dim
        mha_block_flops = dict(
        kqv=3 * h * h  ,
        attention_scores=h * s,
        attn_softmax=SOFTMAX_FLOPS * s * heads,
        attention_dropout=DROPOUT_FLOPS * s * heads,
        attention_scale=s * heads,
        attention_weighted_avg_values=h * s,
        attn_output=h * h,
        attn_output_bias=h,
        attn_output_dropout=DROPOUT_FLOPS * h,
        attn_output_residual=h,
        attn_output_layer_norm=LAYER_NORM_FLOPS * h,)
        ffn_block_flops = dict(
        intermediate=h * i,
        intermediate_act=ACTIVATION_FLOPS * i,
        intermediate_bias=i,
        output=h * i,
        output_bias=h,
        output_dropout=DROPOUT_FLOPS * h,
        output_residual=h,
        output_layer_norm=LAYER_NORM_FLOPS * h,)

        return sum(mha_block_flops.values())*s + sum(ffn_block_flops.values())*s

class MHABlock(nn.Module):
    """
    Multihead Attention block with residual branch
    """
    def __init__(self, dim, num_heads, head_dim=None, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0.,
                 drop_path=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm, group=1, skip_lam=1.):
        super().__init__()
        self.dim = dim
        self.norm1 = norm_layer(dim)
        self.skip_lam = skip_lam
        self.attn = Attention(
            dim, num_heads=num_heads, head_dim=head_dim, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()

    def forward(self, x, padding_mask=None):
        x = x + self.drop_path(self.attn(self.norm1(x*self.skip_lam), padding_mask))/self.skip_lam
        return x

    def flops(self, s):
        heads = self.attn.num_heads
        h = self.dim
        block_flops = dict(
        kqv=3 * h * h ,
        attention_scores=h * s,
        attn_softmax=SOFTMAX_FLOPS * s * heads,
        attention_dropout=DROPOUT_FLOPS * s * heads,
        attention_scale=s * heads,
        attention_weighted_avg_values=h * s,
        attn_output=h * h,
        attn_output_bias=h,
        attn_output_dropout=DROPOUT_FLOPS * h,
        attn_output_residual=h,
        attn_output_layer_norm=LAYER_NORM_FLOPS * h,)

        return sum(block_flops.values())*s

class FFNBlock(nn.Module):
    """
    Feed forward network with residual branch
    """
    def __init__(self, dim, num_heads, head_dim=None, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0.,
                 drop_path=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm, group=1, skip_lam=1.):
        super().__init__()
        self.skip_lam = skip_lam
        self.dim = dim
        self.mlp_hidden_dim = int(dim * mlp_ratio)

        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
        self.norm2 = norm_layer(dim)
        self.mlp = Mlp(in_features=dim, hidden_features=self.mlp_hidden_dim, act_layer=act_layer, drop=drop, group=group)
    def forward(self, x):
        x = x + self.drop_path(self.mlp(self.norm2(x*self.skip_lam)))/self.skip_lam
        return x
    def flops(self, s):
        heads = self.attn.num_heads
        h = self.dim
        i = self.mlp_hidden_dim
        block_flops = dict(
        intermediate=h * i,
        intermediate_act=ACTIVATION_FLOPS * i,
        intermediate_bias=i,
        output=h * i,
        output_bias=h,
        output_dropout=DROPOUT_FLOPS * h,
        output_residual=h,
        output_layer_norm=LAYER_NORM_FLOPS * h,)

        return sum(block_flops.values())*s

class HybridEmbed(nn.Module):
    """ CNN Feature Map Embedding
    Extract feature map from CNN, flatten, project to embedding dim.
    from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
    """
    def __init__(self, backbone, img_size=224, feature_size=None, in_chans=3, embed_dim=768):
        super().__init__()
        assert isinstance(backbone, nn.Module)
        img_size = to_2tuple(img_size)
        self.img_size = img_size
        self.backbone = backbone
        if feature_size is None:
            with torch.no_grad():
                training = backbone.training
                if training:
                    backbone.eval()
                o = self.backbone(torch.zeros(1, in_chans, img_size[0], img_size[1]))[-1]
                feature_size = o.shape[-2:]
                feature_dim = o.shape[1]
                backbone.train(training)
        else:
            feature_size = to_2tuple(feature_size)
            feature_dim = self.backbone.feature_info.channels()[-1]
        self.num_patches = feature_size[0] * feature_size[1]
        self.proj = nn.Conv2d(feature_dim, embed_dim,kernel_size=1)

    def forward(self, x):
        x = self.backbone(x)[-1]
        x = self.proj(x)
        return x


class PatchEmbedNaive(nn.Module):
    """ 
    Image to Patch Embedding
    from https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py
    """
    def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768):
        super().__init__()
        img_size = to_2tuple(img_size)
        patch_size = to_2tuple(patch_size)
        num_patches = (img_size[1] // patch_size[1]) * (img_size[0] // patch_size[0])
        self.img_size = img_size
        self.patch_size = patch_size
        self.num_patches = num_patches
        self.embed_dim = embed_dim

        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)

    def forward(self, x):
        B, C, H, W = x.shape
        assert H == self.img_size[0] and W == self.img_size[1], \
            f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})."
        x = self.proj(x)
        return x

    def flops(self):
        img_size = self.img_size[0]
        block_flops = dict(
        proj=img_size*img_size*3*self.embed_dim,
        )
        return sum(block_flops.values())


class PatchEmbed4_2(nn.Module):
    """ 
    Image to Patch Embedding with 4 layer convolution
    """
    def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768):
        super().__init__()

        new_patch_size = to_2tuple(patch_size // 2)

        img_size = to_2tuple(img_size)
        patch_size = to_2tuple(patch_size)
        num_patches = (img_size[1] // patch_size[1]) * (img_size[0] // patch_size[0])
        self.img_size = img_size
        self.patch_size = patch_size
        self.num_patches = num_patches
        self.embed_dim = embed_dim

        self.conv1 = nn.Conv2d(in_chans, 64, kernel_size=7, stride=2, padding=3, bias=False)  # 112x112
        self.bn1 = nn.BatchNorm2d(64)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1, bias=False)  # 112x112
        self.bn2 = nn.BatchNorm2d(64)
        self.conv3 = nn.Conv2d(64, 64, kernel_size=3, stride=1, padding=1, bias=False)  
        self.bn3 = nn.BatchNorm2d(64)

        self.proj = nn.Conv2d(64, embed_dim, kernel_size=new_patch_size, stride=new_patch_size)
    def forward(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)

        x = self.conv2(x)
        x = self.bn2(x)
        x = self.relu(x)

        x = self.conv3(x)
        x = self.bn3(x)
        x = self.relu(x)

        x = self.proj(x)  # [B, C, W, H]

        return x

    def flops(self):
        img_size = self.img_size[0]
        block_flops = dict(
        conv1=img_size/2*img_size/2*3*64*7*7,
        conv2=img_size/2*img_size/2*64*64*3*3,
        conv3=img_size/2*img_size/2*64*64*3*3,
        proj=img_size/2*img_size/2*64*self.embed_dim,
        )
        return sum(block_flops.values())

    
class PatchEmbed4_2_128(nn.Module):
    """ 
    Image to Patch Embedding with 4 layer convolution and 128 filters
    """
    def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768):
        super().__init__()

        new_patch_size = to_2tuple(patch_size // 2)

        img_size = to_2tuple(img_size)
        patch_size = to_2tuple(patch_size)
        num_patches = (img_size[1] // patch_size[1]) * (img_size[0] // patch_size[0])
        self.img_size = img_size
        self.patch_size = patch_size
        self.num_patches = num_patches
        self.embed_dim = embed_dim

        self.conv1 = nn.Conv2d(in_chans, 128, kernel_size=7, stride=2, padding=3, bias=False)  # 112x112
        self.bn1 = nn.BatchNorm2d(128)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = nn.Conv2d(128, 128, kernel_size=3, stride=1, padding=1, bias=False)  # 112x112
        self.bn2 = nn.BatchNorm2d(128)
        self.conv3 = nn.Conv2d(128, 128, kernel_size=3, stride=1, padding=1, bias=False)  
        self.bn3 = nn.BatchNorm2d(128)

        self.proj = nn.Conv2d(128, embed_dim, kernel_size=new_patch_size, stride=new_patch_size)
    def forward(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)

        x = self.conv2(x)
        x = self.bn2(x)
        x = self.relu(x)

        x = self.conv3(x)
        x = self.bn3(x)
        x = self.relu(x)

        x = self.proj(x)  # [B, C, W, H]

        return x
    def flops(self):
        img_size = self.img_size[0]
        block_flops = dict(
        conv1=img_size/2*img_size/2*3*128*7*7,
        conv2=img_size/2*img_size/2*128*128*3*3,
        conv3=img_size/2*img_size/2*128*128*3*3,
        proj=img_size/2*img_size/2*128*self.embed_dim,
        )
        return sum(block_flops.values())


def get_block(block_type, **kargs):
    if block_type=='mha':
        # multi-head attention block
        return MHABlock(**kargs)
    elif block_type=='ffn':
        # feed forward block
        return FFNBlock(**kargs)
    elif block_type=='tr':
        # transformer block
        return Block(**kargs)


def rand_bbox(size, lam):
    W = size[2]
    H = size[3]
    cut_rat = np.sqrt(1. - lam)
    cut_w = np.int(W * cut_rat)
    cut_h = np.int(H * cut_rat)

    # uniform
    cx = np.random.randint(W)
    cy = np.random.randint(H)

    bbx1 = np.clip(cx - cut_w // 2, 0, W)
    bby1 = np.clip(cy - cut_h // 2, 0, H)
    bbx2 = np.clip(cx + cut_w // 2, 0, W)
    bby2 = np.clip(cy + cut_h // 2, 0, H)

    return bbx1, bby1, bbx2, bby2


def get_dpr(drop_path_rate,depth,drop_path_decay='linear'):
    if drop_path_decay=='linear':
        # linear dpr decay
        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)]  # stochastic depth decay rule
    elif drop_path_decay=='fix':
        # use fixed dpr
        dpr= [drop_path_rate]*depth
    else:
        # use predefined drop_path_rate list
        assert len(drop_path_rate)==depth
        dpr=drop_path_rate
    return dpr

class PredictorLG(nn.Module):
    """ Image to Patch Embedding
    """
    def __init__(self, embed_dim=384):
        super().__init__()
        self.in_conv = nn.Sequential(
            nn.LayerNorm(embed_dim),
            nn.Linear(embed_dim, embed_dim),
            nn.GELU()
        )

        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),
            nn.LogSoftmax(dim=-1)
        )

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

class MultiheadPredictorLG(nn.Module):
    """ Image to Patch Embedding
    """
    def __init__(self, num_heads=6, embed_dim=384):
        super().__init__()

        #print('head_num',num_heads)
        self.num_heads=num_heads
        self.embed_dim = embed_dim
        onehead_in_conv = nn.Sequential(
            nn.LayerNorm(embed_dim // num_heads),
            nn.Linear(embed_dim // num_heads, embed_dim // num_heads),
            nn.GELU()
        )

        onehead_out_conv = nn.Sequential(
            nn.Linear(embed_dim // num_heads, embed_dim // num_heads  // 2),
            nn.GELU(),
            nn.Linear(embed_dim // num_heads // 2, embed_dim // num_heads // 4),
            nn.GELU(),
            nn.Linear(embed_dim // num_heads // 4, 2),
            #nn.LogSoftmax(dim=-1)
        )


        in_conv_list = [onehead_in_conv for _ in range(num_heads)]
        out_conv_list = [onehead_out_conv for _ in range(num_heads)]

        self.in_conv = nn.ModuleList(in_conv_list)
        self.out_conv = nn.ModuleList(out_conv_list)

    def forward(self, x, policy):

        multihead_score = 0
        multihead_softmax_score = 0
        for i in range(self.num_heads):
            x_single = x[:,:,self.embed_dim//self.num_heads*i:self.embed_dim//self.num_heads*(i+1)]   #([96, 196, 64])
            x_single = self.in_conv[i](x_single)
            B, N, C = x_single.size()       #([96, 196, 64])
            local_x = x_single[:,:, :C//2]  #([96, 196, 32])
            global_x = (x_single[:,:, C//2:] * policy).sum(dim=1, keepdim=True) / torch.sum(policy, dim=1, keepdim=True)  #([96, 1, 32])
            x_single = torch.cat([local_x, global_x.expand(B, N, C//2)], dim=-1)  #([96, 196, 64])
            x_single=self.out_conv[i](x_single) #([96, 196, 2])
            
            # for placeholder
            m = nn.Softmax(dim=-1)
            score_softmax = m(x_single)
            multihead_softmax_score += score_softmax

            # for gumble
            n = nn.LogSoftmax(dim=-1)
            score_single = n(x_single)
            multihead_score += score_single

        # for gumble
        multihead_score = multihead_score / self.num_heads  # ([96, 196, 2])

        # for placeholder
        multihead_softmax_score = multihead_softmax_score / self.num_heads  # get softmax keep/drop probability

        return multihead_score, multihead_softmax_score  


class LVViTDiffPruning(nn.Module):
    """ Vision Transformer with tricks
    Arguements:
        p_emb: different conv based position embedding (default: 4 layer conv)
        skip_lam: residual scalar for skip connection (default: 1.0)
        order: which order of layers will be used (default: None, will override depth if given)
        mix_token: use mix token augmentation for batch of tokens (default: False)
        return_dense: whether to return feature of all tokens with an additional aux_head (default: False)
    """
    def __init__(self, img_size=224, patch_size=16, in_chans=3, num_classes=1000, embed_dim=768, depth=12,
                 num_heads=12, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop_rate=0., attn_drop_rate=0.,
                 drop_path_rate=0., drop_path_decay='linear', hybrid_backbone=None, norm_layer=nn.LayerNorm, p_emb='4_2', head_dim = None,
                 skip_lam = 1.0,order=None, mix_token=False, return_dense=False, pruning_loc=None, token_ratio=None, distill=False, viz_mode=False):
        super().__init__()
        self.num_classes = num_classes
        self.num_features = self.embed_dim = embed_dim  # num_features for consistency with other models
        self.output_dim = embed_dim if num_classes==0 else num_classes
        if hybrid_backbone is not None:
            self.patch_embed = HybridEmbed(
                hybrid_backbone, img_size=img_size, in_chans=in_chans, embed_dim=embed_dim)
        else:
            if p_emb=='4_2':
                patch_embed_fn = PatchEmbed4_2
            elif p_emb=='4_2_128':
                patch_embed_fn = PatchEmbed4_2_128
            else:
                patch_embed_fn = PatchEmbedNaive

            self.patch_embed = patch_embed_fn(img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim)

        num_patches = self.patch_embed.num_patches

        self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
        self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + 1, embed_dim))
        self.pos_drop = nn.Dropout(p=drop_rate)

        if order is None:
            dpr=get_dpr(drop_path_rate, depth, drop_path_decay)
            self.blocks = nn.ModuleList([
                Block(
                    dim=embed_dim, num_heads=num_heads, head_dim=head_dim, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
                    drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer, skip_lam=skip_lam)
                for i in range(depth)])
        else:
            # use given order to sequentially generate modules
            dpr=get_dpr(drop_path_rate, len(order), drop_path_decay)
            self.blocks = nn.ModuleList([
                get_block(order[i],
                    dim=embed_dim, num_heads=num_heads, head_dim=head_dim, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
                    drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer, skip_lam=skip_lam)
                for i in range(len(order))])

        self.norm = norm_layer(embed_dim)
        self.head = nn.Linear(embed_dim, num_classes) if num_classes > 0 else nn.Identity()
        
        self.return_dense=return_dense
        self.mix_token=mix_token

        
        predictor_list = [MultiheadPredictorLG(num_heads,embed_dim) for _ in range(len(pruning_loc))]

        self.score_predictor = nn.ModuleList(predictor_list)

        self.pruning_loc = pruning_loc
        self.token_ratio = token_ratio

        if return_dense:
            self.aux_head=nn.Linear(embed_dim, num_classes) if num_classes > 0 else nn.Identity()
        if mix_token:
            self.beta = 1.0
            assert return_dense, "always return all features when mixtoken is enabled"

        self.distill = distill
        self.viz_mode = viz_mode

        trunc_normal_(self.pos_embed, std=.02)
        trunc_normal_(self.cls_token, std=.02)
        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, GroupLinear):
            trunc_normal_(m.group_weight, std=.02)
            if isinstance(m, GroupLinear) and m.group_bias is not None:
                nn.init.constant_(m.group_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 {'pos_embed', 'cls_token'}

    def get_classifier(self):
        return self.head

    def reset_classifier(self, num_classes, global_pool=''):
        self.num_classes = num_classes
        self.head = nn.Linear(self.embed_dim, num_classes) if num_classes > 0 else nn.Identity()

    def forward(self, x):
        x = self.patch_embed(x)
        x = x.flatten(2).transpose(1, 2)
        B = x.shape[0]
        cls_tokens = self.cls_token.expand(B, -1, -1)
        x = torch.cat((cls_tokens, x), dim=1)
        x = x + self.pos_embed
        x = self.pos_drop(x)


        p_count = 0
        out_pred_prob = []
        score_dict = {}
        sparse = []
        init_n = 14 * 14
        prev_decision = torch.ones(B, init_n, 1, dtype=x.dtype, device=x.device)
        policy = torch.ones(B, init_n + 1, 1, dtype=x.dtype, device=x.device)
        if self.viz_mode:
            decisions = [[] for _ in self.pruning_loc]
        for i, blk in enumerate(self.blocks):
            if i in self.pruning_loc:
                spatial_x = x[:, 1:]
                if i != self.pruning_loc[0]:
                    rep_decision = torch.ones(B, p_count, 1, dtype=x.dtype, device=x.device)
                    prev_decision = torch.cat([prev_decision, rep_decision], dim=1)
                pred_score, softmax_score = self.score_predictor[p_count](spatial_x, prev_decision)
                pred_score = pred_score.reshape(B, -1, 2)
                softmax_score = softmax_score.reshape(B, -1, 2)
                #-------------------- 确定 informative token 和 placeholder 的 mask
                if i == self.pruning_loc[0]:
                    hard_keep_decision = F.gumbel_softmax(pred_score, hard=True)[:, :, 0:1] *  prev_decision
                    hard_drop_decision = (1 - hard_keep_decision) - (1 - prev_decision) #  current drop decision
                else:
                    hard_keep_decision_all = F.gumbel_softmax(pred_score, hard=True)[:, :, 0:1] *  prev_decision
                    hard_keep_decision = torch.cat([hard_keep_decision_all[:,:-p_count], rep_decision], dim=1)
                    hard_drop_decision = (1 - hard_keep_decision) - (1 - prev_decision)
                ############### end

                ###get representative token  (regularization)
                softmax_score = softmax_score[:, :, 0:1]  # softmax score of all tokens to keep
                placeholder_score = softmax_score * hard_drop_decision  #keep score of only placeholder tokens
                x2 = spatial_x * placeholder_score  # placehoder score [96, 196, 384]
                x2_sum = torch.sum(x2, dim=1)  # sum by the N dimension, output (B,N,C)-->(B,C) [96, 384]
                x2_sum = torch.unsqueeze(x2_sum, dim=1)  # resize to (B,1,C)  [96, 1, 384]
                #--------------------
                placeholder_score_sum = torch.sum(placeholder_score, dim=1)  # sum of token score, [96, 196, 1]-->[96, 1]
                placeholder_score_sum = torch.unsqueeze(placeholder_score_sum, dim=1)  # resize to [96, 1, 1]
                #--------------------
                represent_token = x2_sum / placeholder_score_sum  # regularization --> [96, 1, 384] representitave token
                rep_mean = represent_token.mean()
                if torch.isnan(rep_mean):
                    print('has nan')
                    represent_token = torch.nan_to_num(represent_token, nan = 1e-6)
             

                x = torch.cat((x,represent_token), dim=1)
                if i != self.pruning_loc[0]:
                    hard_keep_decision = hard_keep_decision[:,:-p_count]

                if self.training:
                    out_pred_prob.append(hard_keep_decision.reshape(B, init_n))
                    cls_policy = torch.ones(B, 1, 1, dtype=hard_keep_decision.dtype, device=hard_keep_decision.device)
                    rep_policy = torch.ones(B, (p_count + 1), 1, dtype=hard_keep_decision.dtype, device=hard_keep_decision.device)
                    policy = torch.cat([cls_policy, hard_keep_decision, rep_policy], dim=1)
                    x = blk(x, policy=policy)
                    prev_decision = hard_keep_decision
                else:
                    cls_policy = torch.ones(B, 1, 1, dtype=hard_keep_decision.dtype, device=hard_keep_decision.device)
                    rep_policy = torch.ones(B, (p_count + 1), 1, dtype=hard_keep_decision.dtype, device=hard_keep_decision.device)
                    policy = torch.cat([cls_policy, hard_keep_decision, rep_policy], dim=1)
                    zeros, unzeros = test_irregular_sparsity(p_count, policy)
                    sparse.append([zeros, unzeros])
                    x = blk(x, policy=policy)
                    prev_decision = hard_keep_decision
                    score = pred_score[:, :, 0:1].cpu().numpy().tolist()
                    score_dict[p_count] = score[0] #144/12=12x30x87x4=125280= 1.5G
                p_count += 1
            else:
                x = blk(x, policy)
        
        x = self.norm(x)
        x_cls = self.head(x[:,0])
        x_aux = self.aux_head(x[:,1:-3])
        final_pred =  x_cls + 0.5 * x_aux.max(1)[0]

        if self.training:
            if self.distill:
                return x_cls, x_aux, prev_decision.detach(), out_pred_prob
            else:
                return final_pred, out_pred_prob
        else:
            with open(file, 'a') as f: # ins
                json.dump(score_dict, f)
                f.write('\n')
            sparse = torch.FloatTensor(sparse).cuda()
            return final_pred, sparse

class LVViT_Teacher(nn.Module):
    """ Vision Transformer with tricks
    Arguements:
        p_emb: different conv based position embedding (default: 4 layer conv)
        skip_lam: residual scalar for skip connection (default: 1.0)
        order: which order of layers will be used (default: None, will override depth if given)
        mix_token: use mix token augmentation for batch of tokens (default: False)
        return_dense: whether to return feature of all tokens with an additional aux_head (default: False)
    """
    def __init__(self, img_size=224, patch_size=16, in_chans=3, num_classes=1000, embed_dim=768, depth=12,
                 num_heads=12, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop_rate=0., attn_drop_rate=0.,
                 drop_path_rate=0., drop_path_decay='linear', hybrid_backbone=None, norm_layer=nn.LayerNorm, p_emb='4_2', head_dim = None,
                 skip_lam = 1.0,order=None, mix_token=False, return_dense=False):
        super().__init__()
        self.num_classes = num_classes
        self.num_features = self.embed_dim = embed_dim  # num_features for consistency with other models
        self.output_dim = embed_dim if num_classes==0 else num_classes
        if hybrid_backbone is not None:
            self.patch_embed = HybridEmbed(
                hybrid_backbone, img_size=img_size, in_chans=in_chans, embed_dim=embed_dim)
        else:
            if p_emb=='4_2':
                patch_embed_fn = PatchEmbed4_2
            elif p_emb=='4_2_128':
                patch_embed_fn = PatchEmbed4_2_128
            else:
                patch_embed_fn = PatchEmbedNaive

            self.patch_embed = patch_embed_fn(img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim)

        num_patches = self.patch_embed.num_patches

        self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
        self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + 1, embed_dim))
        self.pos_drop = nn.Dropout(p=drop_rate)

        if order is None:
            dpr=get_dpr(drop_path_rate, depth, drop_path_decay)
            self.blocks = nn.ModuleList([
                Block(
                    dim=embed_dim, num_heads=num_heads, head_dim=head_dim, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
                    drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer, skip_lam=skip_lam)
                for i in range(depth)])
        else:
            # use given order to sequentially generate modules
            dpr=get_dpr(drop_path_rate, len(order), drop_path_decay)
            self.blocks = nn.ModuleList([
                get_block(order[i],
                    dim=embed_dim, num_heads=num_heads, head_dim=head_dim, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
                    drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer, skip_lam=skip_lam)
                for i in range(len(order))])

        self.norm = norm_layer(embed_dim)
        self.head = nn.Linear(embed_dim, num_classes) if num_classes > 0 else nn.Identity()
        
        self.return_dense=return_dense
        self.mix_token=mix_token

        if return_dense:
            self.aux_head=nn.Linear(embed_dim, num_classes) if num_classes > 0 else nn.Identity()
        if mix_token:
            self.beta = 1.0
            assert return_dense, "always return all features when mixtoken is enabled"

        trunc_normal_(self.pos_embed, std=.02)
        trunc_normal_(self.cls_token, std=.02)
        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, GroupLinear):
            trunc_normal_(m.group_weight, std=.02)
            if isinstance(m, GroupLinear) and m.group_bias is not None:
                nn.init.constant_(m.group_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 {'pos_embed', 'cls_token'}

    def get_classifier(self):
        return self.head

    def reset_classifier(self, num_classes, global_pool=''):
        self.num_classes = num_classes
        self.head = nn.Linear(self.embed_dim, num_classes) if num_classes > 0 else nn.Identity()

    def forward(self, x):
        x = self.patch_embed(x)
        x = x.flatten(2).transpose(1, 2)
        B = x.shape[0]
        cls_tokens = self.cls_token.expand(B, -1, -1)
        x = torch.cat((cls_tokens, x), dim=1)
        x = x + self.pos_embed
        x = self.pos_drop(x)

        for i, blk in enumerate(self.blocks):
            x = blk(x)
        
        x = self.norm(x)
        x_cls = self.head(x[:,0])
        x_aux = self.aux_head(x[:,1:])
        return x_cls, x_aux


def test_irregular_sparsity(name,matrix):

    # continue
    zeros = np.sum(matrix.cpu().detach().numpy() == 0)

    non_zeros = np.sum(matrix.cpu().detach().numpy() != 0)

    # print(name, non_zeros)
    #print(" {}, all weights: {}, irregular zeros: {}, irregular sparsity is: {:.4f}".format( name, zeros+non_zeros, zeros, zeros / (zeros + non_zeros)))
    # print(non_zeros+zeros)
    # total_nonzeros += 128000

    return zeros,non_zeros