vit.py

""" Vision Transformer (ViT) in PyTorch

A PyTorch implement of Vision Transformers as described in
'An Image Is Worth 16 x 16 Words: Transformers for Image Recognition at Scale' - https://arxiv.org/abs/2010.11929

The official jax code is released and available at https://github.com/google-research/vision_transformer

Acknowledgments:
* The paper authors for releasing code and weights, thanks!
* I fixed my class token impl based on Phil Wang's https://github.com/lucidrains/vit-pytorch ... check it out
for some einops/einsum fun
* Simple transformer style inspired by Andrej Karpathy's https://github.com/karpathy/minGPT
* Bert reference code checks against Huggingface Transformers and Tensorflow Bert

DeiT model defs and weights from https://github.com/facebookresearch/deit,
paper `DeiT: Data-efficient Image Transformers` - https://arxiv.org/abs/2012.12877

Hacked together by / Copyright 2020 Ross Wightman
"""
import math
import logging
from functools import partial
from collections import OrderedDict
from copy import Error, deepcopy
from sre_constants import AT_NON_BOUNDARY

import torch
import torch.nn as nn
import torch.nn.functional as F

from utils import batch_index_select

from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
from timm.models.layers import DropPath, to_2tuple, trunc_normal_

_logger = logging.getLogger(__name__)


def _cfg(url='', **kwargs):
    return {
        'url': url,
        'num_classes': 1000, 'input_size': (3, 224, 224), 'pool_size': None,
        'crop_pct': .9, 'interpolation': 'bicubic',
        'mean': IMAGENET_DEFAULT_MEAN, 'std': IMAGENET_DEFAULT_STD,
        'first_conv': 'patch_embed.proj', 'classifier': 'head',
        **kwargs
    }


default_cfgs = {
    # patch models (my experiments)
    'vit_small_patch16_224': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/vit_small_p16_224-15ec54c9.pth',
    ),

    # patch models (weights ported from official Google JAX impl)
    'vit_base_patch16_224': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_p16_224-80ecf9dd.pth',
        mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5),
    ),
    'vit_base_patch32_224': _cfg(
        url='',  # no official model weights for this combo, only for in21k
        mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
    'vit_base_patch16_384': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_p16_384-83fb41ba.pth',
        input_size=(3, 384, 384), mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5), crop_pct=1.0),
    'vit_base_patch32_384': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_p32_384-830016f5.pth',
        input_size=(3, 384, 384), mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5), crop_pct=1.0),
    'vit_large_patch16_224': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_large_p16_224-4ee7a4dc.pth',
        mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
    'vit_large_patch32_224': _cfg(
        url='',  # no official model weights for this combo, only for in21k
        mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
    'vit_large_patch16_384': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_large_p16_384-b3be5167.pth',
        input_size=(3, 384, 384), mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5), crop_pct=1.0),
    'vit_large_patch32_384': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_large_p32_384-9b920ba8.pth',
        input_size=(3, 384, 384), mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5), crop_pct=1.0),

    # patch models, imagenet21k (weights ported from official Google JAX impl)
    'vit_base_patch16_224_in21k': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_patch16_224_in21k-e5005f0a.pth',
        num_classes=21843, mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
    'vit_base_patch32_224_in21k': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_patch32_224_in21k-8db57226.pth',
        num_classes=21843, mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
    'vit_large_patch16_224_in21k': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_large_patch16_224_in21k-606da67d.pth',
        num_classes=21843, mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
    'vit_large_patch32_224_in21k': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_large_patch32_224_in21k-9046d2e7.pth',
        num_classes=21843, mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
    'vit_huge_patch14_224_in21k': _cfg(
        hf_hub='timm/vit_huge_patch14_224_in21k',
        num_classes=21843, mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),

    # hybrid models (weights ported from official Google JAX impl)
    'vit_base_resnet50_224_in21k': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_resnet50_224_in21k-6f7c7740.pth',
        num_classes=21843, mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5), crop_pct=0.9, first_conv='patch_embed.backbone.stem.conv'),
    'vit_base_resnet50_384': _cfg(
        url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_resnet50_384-9fd3c705.pth',
        input_size=(3, 384, 384), mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5), crop_pct=1.0, first_conv='patch_embed.backbone.stem.conv'),

    # hybrid models (my experiments)
    'vit_small_resnet26d_224': _cfg(),
    'vit_small_resnet50d_s3_224': _cfg(),
    'vit_base_resnet26d_224': _cfg(),
    'vit_base_resnet50d_224': _cfg(),

    # deit models (FB weights)
    'vit_deit_tiny_patch16_224': _cfg(
        url='https://dl.fbaipublicfiles.com/deit/deit_tiny_patch16_224-a1311bcf.pth'),
    'vit_deit_small_patch16_224': _cfg(
        url='https://dl.fbaipublicfiles.com/deit/deit_small_patch16_224-cd65a155.pth'),
    'vit_deit_base_patch16_224': _cfg(
        url='https://dl.fbaipublicfiles.com/deit/deit_base_patch16_224-b5f2ef4d.pth',),
    'vit_deit_base_patch16_384': _cfg(
        url='https://dl.fbaipublicfiles.com/deit/deit_base_patch16_384-8de9b5d1.pth',
        input_size=(3, 384, 384), crop_pct=1.0),
    'vit_deit_tiny_distilled_patch16_224': _cfg(
        url='https://dl.fbaipublicfiles.com/deit/deit_tiny_distilled_patch16_224-b40b3cf7.pth'),
    'vit_deit_small_distilled_patch16_224': _cfg(
        url='https://dl.fbaipublicfiles.com/deit/deit_small_distilled_patch16_224-649709d9.pth'),
    'vit_deit_base_distilled_patch16_224': _cfg(
        url='https://dl.fbaipublicfiles.com/deit/deit_base_distilled_patch16_224-df68dfff.pth', ),
    'vit_deit_base_distilled_patch16_384': _cfg(
        url='https://dl.fbaipublicfiles.com/deit/deit_base_distilled_patch16_384-d0272ac0.pth',
        input_size=(3, 384, 384), crop_pct=1.0),
}


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)

    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 Attention(nn.Module):
    def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0.):
        super().__init__()
        self.num_heads = num_heads
        head_dim = dim // num_heads
        # NOTE scale factor was wrong in my original version, can set manually to be compat with prev weights
        self.scale = qk_scale or head_dim ** -0.5

        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(dim, 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
        # 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):
        B, N, C = x.shape
        qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
        q, k, v = qkv[0], qkv[1], qkv[2]   # make torchscript happy (cannot use tensor as tuple)

        attn = (q @ k.transpose(-2, -1)) * self.scale

        if policy is None:
            attn = attn.softmax(dim=-1)
        else:
            attn = self.softmax_with_policy(attn, policy)

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

        # token-level head importance
        cls_attn = attn[:, :, 0, 1:]  # [B, H, N-1]
        head_importance = attn_out[..., 1:, :].norm(dim=-1)
        head_importance = head_importance / (head_importance.sum(dim=1, keepdim=True) + 1e-8)
        cls_attn = cls_attn * head_importance
        cls_attn = cls_attn.sum(dim=1)  # [B, N-1]
        
        return x, cls_attn


class AttentionT(nn.Module):
    def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0.):
        super().__init__()
        self.num_heads = num_heads
        head_dim = dim // num_heads
        # NOTE scale factor was wrong in my original version, can set manually to be compat with prev weights
        self.scale = qk_scale or head_dim ** -0.5

        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
        self.attn_drop = nn.Dropout(attn_drop)
        self.proj = nn.Linear(dim, 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):
        B, N, C = x.shape
        qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
        q, k, v = qkv[0], qkv[1], qkv[2]   # make torchscript happy (cannot use tensor as tuple)

        attn = (q @ k.transpose(-2, -1)) * self.scale

        if policy is None:
            attn = attn.softmax(dim=-1)
        else:
            attn = self.softmax_with_policy(attn, policy)

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


class Block(nn.Module):

    def __init__(self, dim, num_heads, 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):
        super().__init__()
        self.norm1 = norm_layer(dim)
        self.dim = dim
        self.mlp_dim = dim * mlp_ratio
        self.attn = Attention(
            dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
        # NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
        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)
        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)

    def forward(self, x, policy=None):
        tmp, attn = self.attn(self.norm1(x), policy=policy)
        x = x + self.drop_path(tmp)
        x = x + self.drop_path(self.mlp(self.norm2(x)))

        if self.training:
            flops = self.flops(policy.squeeze(dim=-1).sum(dim=-1))
            return x, attn, flops
        else:
            return x, attn

    def flops(self, num_tokens):
        dim = self.dim
        mlp_dim = self.mlp_dim
        # mha_block_flops
        kqv_flops = 3 * dim * dim * num_tokens
        attn_flops = dim * num_tokens * num_tokens
        attnv_flops = dim * num_tokens * num_tokens
        attnout_flops = dim * dim * num_tokens
        mha_block_flops = kqv_flops + attn_flops + attnv_flops + attnout_flops
        # ffn_flops
        fc1_flops = dim * mlp_dim * num_tokens
        fc2_flops = dim * num_tokens * mlp_dim
        ffn_flops = fc1_flops + fc2_flops
        return (mha_block_flops + ffn_flops) / 1e+9


class BlockT(nn.Module):

    def __init__(self, dim, num_heads, 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):
        super().__init__()
        self.norm1 = norm_layer(dim)
        self.dim = dim
        self.mlp_dim = dim * mlp_ratio
        self.attn = AttentionT(
            dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop)
        # NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
        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)
        self.mlp = Mlp(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=drop)

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

    def flops(self, num_tokens):
        dim = self.dim
        mlp_dim = self.mlp_dim
        # mha_block_flops
        kqv_flops = 3 * dim * dim * num_tokens
        attn_flops = dim * num_tokens * num_tokens
        attnv_flops = dim * num_tokens * num_tokens
        attnout_flops = dim * dim * num_tokens
        mha_block_flops = kqv_flops + attn_flops + attnv_flops + attnout_flops
        # ffn_flops
        fc1_flops = dim * mlp_dim * num_tokens
        fc2_flops = dim * num_tokens * mlp_dim
        ffn_flops = fc1_flops + fc2_flops
        return (mha_block_flops + ffn_flops) / 1e+9


class PatchEmbed(nn.Module):
    """ Image to Patch Embedding
    """
    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
        # FIXME look at relaxing size constraints
        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).flatten(2).transpose(1, 2)
        return x
    
    def flops(self):
        patch_size = self.patch_size[0]
        block_flops = self.num_patches * self.embed_dim * 3 * patch_size * patch_size
        return block_flops / 1e+9


class HybridEmbed(nn.Module):
    """ CNN Feature Map Embedding
    Extract feature map from CNN, flatten, project to embedding dim.
    """
    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():
                # FIXME this is hacky, but most reliable way of determining the exact dim of the output feature
                # map for all networks, the feature metadata has reliable channel and stride info, but using
                # stride to calc feature dim requires info about padding of each stage that isn't captured.
                training = backbone.training
                if training:
                    backbone.eval()
                o = self.backbone(torch.zeros(1, in_chans, img_size[0], img_size[1]))
                if isinstance(o, (list, tuple)):
                    o = o[-1]  # last feature if backbone outputs list/tuple of features
                feature_size = o.shape[-2:]
                feature_dim = o.shape[1]
                backbone.train(training)
        else:
            feature_size = to_2tuple(feature_size)
            if hasattr(self.backbone, 'feature_info'):
                feature_dim = self.backbone.feature_info.channels()[-1]
            else:
                feature_dim = self.backbone.num_features
        self.num_patches = feature_size[0] * feature_size[1]
        self.proj = nn.Conv2d(feature_dim, embed_dim, 1)

    def forward(self, x):
        x = self.backbone(x)
        if isinstance(x, (list, tuple)):
            x = x[-1]  # last feature if backbone outputs list/tuple of features
        x = self.proj(x).flatten(2).transpose(1, 2)
        return x


class AdaptiveSparsityModule(nn.Module):
    def __init__(self, threshold_init):
        super().__init__()
        self.keep_threshold = nn.Parameter(torch.tensor(threshold_init))

    def forward(self, score, tau=1):
        B = score.size(0)
        if not self.training:
            idx = score > self.keep_threshold
            if B==1:
                return score, idx
            k = idx.sum().item() // B
            return score, k
        
        y_soft = torch.sigmoid((score - self.keep_threshold) / tau)
        y_hard = (y_soft > 0.5).float()
        ret = y_hard - y_soft.detach() + y_soft
        return score, ret


class VisionTransformerDiffPruning(nn.Module):
    """ Vision Transformer

    A PyTorch impl of : `An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale`  -
        https://arxiv.org/abs/2010.11929
    """
    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=True, qk_scale=None, representation_size=None,
                 drop_rate=0., attn_drop_rate=0., drop_path_rate=0., hybrid_backbone=None, norm_layer=None, 
                 pruning_loc=None, token_ratio=None, accu=False, ema_accu=False, distill=False, viz_mode=False):
        """
        Args:
            img_size (int, tuple): input image size
            patch_size (int, tuple): patch size
            in_chans (int): number of input channels
            num_classes (int): number of classes for classification head
            embed_dim (int): embedding dimension
            depth (int): depth of transformer
            num_heads (int): number of attention heads
            mlp_ratio (int): ratio of mlp hidden dim to embedding dim
            qkv_bias (bool): enable bias for qkv if True
            qk_scale (float): override default qk scale of head_dim ** -0.5 if set
            representation_size (Optional[int]): enable and set representation layer (pre-logits) to this value if set
            drop_rate (float): dropout rate
            attn_drop_rate (float): attention dropout rate
            drop_path_rate (float): stochastic depth rate
            hybrid_backbone (nn.Module): CNN backbone to use in-place of PatchEmbed module
            norm_layer: (nn.Module): normalization layer
        """
        super().__init__()
        self.img_size = img_size
        self.beta = 0.99
        self.target_index = [2,3,4,5,6,7,8,9,10,11]
        self.layer_depth = depth

        print('## diff vit pruning method')
        self.num_classes = num_classes
        self.num_features = self.embed_dim = embed_dim  # num_features for consistency with other models
        norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6)

        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:
            self.patch_embed = PatchEmbed(
                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)

        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)]  # stochastic depth decay rule
        self.blocks = nn.ModuleList([
            Block(
                dim=embed_dim, num_heads=num_heads, 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)
            for i in range(depth)])
        self.norm = norm_layer(embed_dim)

        # Representation layer
        if representation_size:
            self.num_features = representation_size
            self.pre_logits = nn.Sequential(OrderedDict([
                ('fc', nn.Linear(embed_dim, representation_size)),
                ('act', nn.Tanh())
            ]))
        else:
            self.pre_logits = nn.Identity()

        # Classifier head
        self.head = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()

        self.accu = accu
        self.ema_accu = ema_accu
        if accu:
            threshold_init = [0.004, 0.012, 0.02]
        else:
            threshold_init = [0.001, 0.002, 0.003]

        print('## threshold_init: ', end='')
        print(threshold_init)
        predictor_list = [AdaptiveSparsityModule(threshold_init[i]) for i in range(len(pruning_loc))]

        self.score_predictor = nn.ModuleList(predictor_list)

        self.distill = distill
        self.viz_mode = viz_mode

        self.pruning_loc = pruning_loc
        self.token_ratio = token_ratio

        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, nn.LayerNorm):
            nn.init.constant_(m.bias, 0)
            nn.init.constant_(m.weight, 1.0)

    @torch.jit.ignore
    def no_weight_decay(self):
        d = {'pos_embed', 'cls_token'}
        for i in range(len(self.pruning_loc)):
            d.add('score_predictor.{}.keep_threshold'.format(i))
        return d

    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):
        B = x.shape[0]
        x = self.patch_embed(x)

        cls_tokens = self.cls_token.expand(B, -1, -1)  # stole cls_tokens impl from Phil Wang, thanks
        x = torch.cat((cls_tokens, x), dim=1)
        x = x + self.pos_embed
        x = self.pos_drop(x)

        p_count = 0
        total_flops = self.patch_embed.flops()
        global_cls_attn = 0
        out_pred_prob = []
        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:
                if self.training:
                    score, hard_keep_decision = self.score_predictor[p_count](global_cls_attn, tau=1e-4)
                    hard_keep_decision = hard_keep_decision[:, :, None]
                    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)
                    policy = torch.cat([cls_policy, hard_keep_decision], dim=1)
                    x, attn, flops = blk(x, policy=policy)
                    prev_decision = hard_keep_decision
                    total_flops += flops
                else:
                    if B==1:
                        score, mask = self.score_predictor[p_count](global_cls_attn, tau=1)
                        keep_policy = torch.nonzero(mask.squeeze()).squeeze().unsqueeze(dim=0)
                    else:
                        score, remain_num = self.score_predictor[p_count](global_cls_attn, tau=1)
                        num_keep_node = remain_num
                        keep_policy = torch.argsort(score, dim=1, descending=True)[:, :num_keep_node]
                    if self.viz_mode:
                        decisions[p_count].append(keep_policy)
                    cls_policy = torch.zeros(B, 1, dtype=keep_policy.dtype, device=keep_policy.device)
                    now_policy = torch.cat([cls_policy, keep_policy + 1], dim=1)
                    x = batch_index_select(x, now_policy)
                    prev_decision = batch_index_select(prev_decision, keep_policy)
                    x, attn = blk(x)
                    global_cls_attn = batch_index_select(global_cls_attn, keep_policy)
                p_count += 1
            else:
                if self.training:
                    x, attn, flops = blk(x, policy)
                    total_flops += flops
                else:
                    x, attn = blk(x)

            if i in self.target_index:
                if self.accu:
                    global_cls_attn += attn
                elif self.ema_accu:
                    if i==self.target_index[0]:
                        global_cls_attn = attn
                    else:
                        global_cls_attn = self.beta*global_cls_attn + (1-self.beta)*attn
                else:
                    global_cls_attn = attn

        x = self.norm(x)
        features = x[:, 1:]
        x = x[:, 0]
        x = self.pre_logits(x)
        x = self.head(x)

        if self.training:
            if self.distill:
                return x, features, prev_decision.detach(), out_pred_prob, total_flops, (flops_1stage, flops_2stage, flops_3stage, flops_4stage)
            else:
                return x, out_pred_prob
        else:
            if self.viz_mode:
                return x, decisions
            else:
                return x


class VisionTransformerTeacher(nn.Module):
    """ Vision Transformer

    A PyTorch impl of : `An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale`  -
        https://arxiv.org/abs/2010.11929
    """
    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=True, qk_scale=None, representation_size=None,
                 drop_rate=0., attn_drop_rate=0., drop_path_rate=0., hybrid_backbone=None, norm_layer=None):
        """
        Args:
            img_size (int, tuple): input image size
            patch_size (int, tuple): patch size
            in_chans (int): number of input channels
            num_classes (int): number of classes for classification head
            embed_dim (int): embedding dimension
            depth (int): depth of transformer
            num_heads (int): number of attention heads
            mlp_ratio (int): ratio of mlp hidden dim to embedding dim
            qkv_bias (bool): enable bias for qkv if True
            qk_scale (float): override default qk scale of head_dim ** -0.5 if set
            representation_size (Optional[int]): enable and set representation layer (pre-logits) to this value if set
            drop_rate (float): dropout rate
            attn_drop_rate (float): attention dropout rate
            drop_path_rate (float): stochastic depth rate
            hybrid_backbone (nn.Module): CNN backbone to use in-place of PatchEmbed module
            norm_layer: (nn.Module): normalization layer
        """
        super().__init__()

        print('## diff vit pruning method')
        self.num_classes = num_classes
        self.num_features = self.embed_dim = embed_dim  # num_features for consistency with other models
        norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6)

        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:
            self.patch_embed = PatchEmbed(
                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)

        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)]  # stochastic depth decay rule
        self.blocks = nn.ModuleList([
            BlockT(
                dim=embed_dim, num_heads=num_heads, 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)
            for i in range(depth)])
        self.norm = norm_layer(embed_dim)

        # Representation layer
        if representation_size:
            self.num_features = representation_size
            self.pre_logits = nn.Sequential(OrderedDict([
                ('fc', nn.Linear(embed_dim, representation_size)),
                ('act', nn.Tanh())
            ]))
        else:
            self.pre_logits = nn.Identity()

        # Classifier head
        self.head = nn.Linear(self.num_features, num_classes) if num_classes > 0 else nn.Identity()

        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, 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):
        B = x.shape[0]
        x = self.patch_embed(x)

        cls_tokens = self.cls_token.expand(B, -1, -1)  # stole cls_tokens impl from Phil Wang, thanks
        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)

        feature = self.norm(x)
        cls = feature[:, 0]
        tokens = feature[:, 1:]
        cls = self.pre_logits(cls)
        cls = self.head(cls)
        return cls, tokens

    def cal_flops(self):
        num_tokens = self.patch_embed.num_patches
        flops = self.patch_embed.flops()
        for i, blk in enumerate(self.blocks):
            flops += blk.flops(num_tokens)
        return flops


def resize_pos_embed(posemb, posemb_new):
    # Rescale the grid of position embeddings when loading from state_dict. Adapted from
    # https://github.com/google-research/vision_transformer/blob/00883dd691c63a6830751563748663526e811cee/vit_jax/checkpoint.py#L224
    _logger.info('Resized position embedding: %s to %s', posemb.shape, posemb_new.shape)
    ntok_new = posemb_new.shape[1]
    if True:
        posemb_tok, posemb_grid = posemb[:, :1], posemb[0, 1:]
        ntok_new -= 1
    else:
        posemb_tok, posemb_grid = posemb[:, :0], posemb[0]
    gs_old = int(math.sqrt(len(posemb_grid)))
    gs_new = int(math.sqrt(ntok_new))
    _logger.info('Position embedding grid-size from %s to %s', gs_old, gs_new)
    posemb_grid = posemb_grid.reshape(1, gs_old, gs_old, -1).permute(0, 3, 1, 2)
    posemb_grid = F.interpolate(posemb_grid, size=(gs_new, gs_new), mode='bilinear')
    posemb_grid = posemb_grid.permute(0, 2, 3, 1).reshape(1, gs_new * gs_new, -1)
    posemb = torch.cat([posemb_tok, posemb_grid], dim=1)
    return posemb


def checkpoint_filter_fn(state_dict, model):
    """ convert patch embedding weight from manual patchify + linear proj to conv"""
    out_dict = {}
    if 'model' in state_dict:
        # For deit models
        state_dict = state_dict['model']
    for k, v in state_dict.items():
        if 'patch_embed.proj.weight' in k and len(v.shape) < 4:
            # For old models that I trained prior to conv based patchification
            O, I, H, W = model.patch_embed.proj.weight.shape
            v = v.reshape(O, -1, H, W)
        elif k == 'pos_embed' and v.shape != model.pos_embed.shape:
            # To resize pos embedding when using model at different size from pretrained weights
            v = resize_pos_embed(v, model.pos_embed)
        out_dict[k] = v
    return out_dict