|
@@ -0,0 +1,671 @@
|
|
|
|
|
+""" 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
|
|
|
|
|
+
|
|
|
|
|
+Status/TODO:
|
|
|
|
|
+* Models updated to be compatible with official impl. Args added to support backward compat for old PyTorch weights.
|
|
|
|
|
+* Weights ported from official jax impl for 384x384 base and small models, 16x16 and 32x32 patches.
|
|
|
|
|
+* Trained (supervised on ImageNet-1k) my custom 'small' patch model to 77.9, 'base' to 79.4 top-1 with this code.
|
|
|
|
|
+* Hopefully find time and GPUs for SSL or unsupervised pretraining on OpenImages w/ ImageNet fine-tune in future.
|
|
|
|
|
+
|
|
|
|
|
+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
|
|
|
|
|
+
|
|
|
|
|
+Hacked together by / Copyright 2020 Ross Wightman
|
|
|
|
|
+"""
|
|
|
|
|
+import warnings
|
|
|
|
|
+import math
|
|
|
|
|
+import torch
|
|
|
|
|
+from functools import partial
|
|
|
|
|
+import torch.nn as nn
|
|
|
|
|
+import torch.nn.functional as F
|
|
|
|
|
+import torch.utils.checkpoint as checkpoint
|
|
|
|
|
+from timm.models.layers import drop_path, to_2tuple, trunc_normal_
|
|
|
|
|
+
|
|
|
|
|
+
|
|
|
|
|
+def _cfg(url='', **kwargs):
|
|
|
|
|
+ return {
|
|
|
|
|
+ 'url': url,
|
|
|
|
|
+ 'num_classes': 1000, 'input_size': (3, 224, 224), 'pool_size': None,
|
|
|
|
|
+ 'crop_pct': .9, 'interpolation': 'bicubic',
|
|
|
|
|
+ 'mean': (0.5, 0.5, 0.5), 'std': (0.5, 0.5, 0.5),
|
|
|
|
|
+ **kwargs
|
|
|
|
|
+ }
|
|
|
|
|
+
|
|
|
|
|
+
|
|
|
|
|
+class DropPath(nn.Module):
|
|
|
|
|
+ """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
|
|
|
|
|
+ """
|
|
|
|
|
+
|
|
|
|
|
+ def __init__(self, drop_prob=None):
|
|
|
|
|
+ super(DropPath, self).__init__()
|
|
|
|
|
+ self.drop_prob = drop_prob
|
|
|
|
|
+
|
|
|
|
|
+ def forward(self, x):
|
|
|
|
|
+ return drop_path(x, self.drop_prob, self.training)
|
|
|
|
|
+
|
|
|
|
|
+ def extra_repr(self) -> str:
|
|
|
|
|
+ return 'p={}'.format(self.drop_prob)
|
|
|
|
|
+
|
|
|
|
|
+
|
|
|
|
|
+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)
|
|
|
|
|
+ # commit this for the orignal BERT implement
|
|
|
|
|
+ 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., window_size=None, attn_head_dim=None):
|
|
|
|
|
+ super().__init__()
|
|
|
|
|
+ self.num_heads = num_heads
|
|
|
|
|
+ head_dim = dim // num_heads
|
|
|
|
|
+ if attn_head_dim is not None:
|
|
|
|
|
+ head_dim = attn_head_dim
|
|
|
|
|
+ all_head_dim = head_dim * self.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, all_head_dim * 3, bias=False)
|
|
|
|
|
+ if qkv_bias:
|
|
|
|
|
+ self.q_bias = nn.Parameter(torch.zeros(all_head_dim))
|
|
|
|
|
+ self.v_bias = nn.Parameter(torch.zeros(all_head_dim))
|
|
|
|
|
+ else:
|
|
|
|
|
+ self.q_bias = None
|
|
|
|
|
+ self.v_bias = None
|
|
|
|
|
+
|
|
|
|
|
+ if window_size:
|
|
|
|
|
+ self.window_size = window_size
|
|
|
|
|
+ self.num_relative_distance = (2 * window_size[0] - 1) * (2 * window_size[1] - 1) + 3
|
|
|
|
|
+ self.relative_position_bias_table = nn.Parameter(
|
|
|
|
|
+ torch.zeros(self.num_relative_distance, num_heads)) # 2*Wh-1 * 2*Ww-1, nH
|
|
|
|
|
+ # cls to token & token 2 cls & cls to cls
|
|
|
|
|
+
|
|
|
|
|
+ # get pair-wise relative position index for each token inside the window
|
|
|
|
|
+ coords_h = torch.arange(window_size[0])
|
|
|
|
|
+ coords_w = torch.arange(window_size[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] += window_size[0] - 1 # shift to start from 0
|
|
|
|
|
+ relative_coords[:, :, 1] += window_size[1] - 1
|
|
|
|
|
+ relative_coords[:, :, 0] *= 2 * window_size[1] - 1
|
|
|
|
|
+ relative_position_index = \
|
|
|
|
|
+ torch.zeros(size=(window_size[0] * window_size[1] + 1,) * 2, dtype=relative_coords.dtype)
|
|
|
|
|
+ relative_position_index[1:, 1:] = relative_coords.sum(-1) # Wh*Ww, Wh*Ww
|
|
|
|
|
+ relative_position_index[0, 0:] = self.num_relative_distance - 3
|
|
|
|
|
+ relative_position_index[0:, 0] = self.num_relative_distance - 2
|
|
|
|
|
+ relative_position_index[0, 0] = self.num_relative_distance - 1
|
|
|
|
|
+
|
|
|
|
|
+ self.register_buffer("relative_position_index", relative_position_index)
|
|
|
|
|
+
|
|
|
|
|
+ # trunc_normal_(self.relative_position_bias_table, std=.0)
|
|
|
|
|
+ else:
|
|
|
|
|
+ self.window_size = None
|
|
|
|
|
+ self.relative_position_bias_table = None
|
|
|
|
|
+ self.relative_position_index = None
|
|
|
|
|
+
|
|
|
|
|
+ self.attn_drop = nn.Dropout(attn_drop)
|
|
|
|
|
+ self.proj = nn.Linear(all_head_dim, dim)
|
|
|
|
|
+ self.proj_drop = nn.Dropout(proj_drop)
|
|
|
|
|
+
|
|
|
|
|
+ def forward(self, x, rel_pos_bias=None, training_window_size=None):
|
|
|
|
|
+ B, N, C = x.shape
|
|
|
|
|
+ qkv_bias = None
|
|
|
|
|
+ if self.q_bias is not None:
|
|
|
|
|
+ qkv_bias = torch.cat((self.q_bias, torch.zeros_like(self.v_bias, requires_grad=False), self.v_bias))
|
|
|
|
|
+ # qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
|
|
|
|
|
+ qkv = F.linear(input=x, weight=self.qkv.weight, bias=qkv_bias)
|
|
|
|
|
+ qkv = qkv.reshape(B, N, 3, self.num_heads, -1).permute(2, 0, 3, 1, 4)
|
|
|
|
|
+ q, k, v = qkv[0], qkv[1], qkv[2] # make torchscript happy (cannot use tensor as tuple)
|
|
|
|
|
+
|
|
|
|
|
+ q = q * self.scale
|
|
|
|
|
+ attn = (q @ k.transpose(-2, -1))
|
|
|
|
|
+
|
|
|
|
|
+ if self.relative_position_bias_table is not None:
|
|
|
|
|
+ if training_window_size == self.window_size:
|
|
|
|
|
+ relative_position_bias = \
|
|
|
|
|
+ self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
|
|
|
|
|
+ self.window_size[0] * self.window_size[1] + 1,
|
|
|
|
|
+ self.window_size[0] * self.window_size[1] + 1, -1) # Wh*Ww,Wh*Ww,nH
|
|
|
|
|
+ relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww
|
|
|
|
|
+ attn = attn + relative_position_bias.unsqueeze(0)
|
|
|
|
|
+ else:
|
|
|
|
|
+ training_window_size = tuple(training_window_size.tolist())
|
|
|
|
|
+ new_num_relative_distance = (2 * training_window_size[0] - 1) * (2 * training_window_size[1] - 1) + 3
|
|
|
|
|
+ # new_num_relative_dis 为 所有可能的相对位置选项,包含cls-cls,tok-cls,与cls-tok
|
|
|
|
|
+ new_relative_position_bias_table = F.interpolate(
|
|
|
|
|
+ self.relative_position_bias_table[:-3, :].permute(1, 0).view(1, self.num_heads,
|
|
|
|
|
+ 2 * self.window_size[0] - 1,
|
|
|
|
|
+ 2 * self.window_size[1] - 1),
|
|
|
|
|
+ size=(2 * training_window_size[0] - 1, 2 * training_window_size[1] - 1), mode='bicubic',
|
|
|
|
|
+ align_corners=False)
|
|
|
|
|
+ new_relative_position_bias_table = new_relative_position_bias_table.view(self.num_heads,
|
|
|
|
|
+ new_num_relative_distance - 3).permute(
|
|
|
|
|
+ 1, 0)
|
|
|
|
|
+ new_relative_position_bias_table = torch.cat(
|
|
|
|
|
+ [new_relative_position_bias_table, self.relative_position_bias_table[-3::]], dim=0)
|
|
|
|
|
+
|
|
|
|
|
+ # get pair-wise relative position index for each token inside the window
|
|
|
|
|
+ coords_h = torch.arange(training_window_size[0])
|
|
|
|
|
+ coords_w = torch.arange(training_window_size[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] += training_window_size[0] - 1 # shift to start from 0
|
|
|
|
|
+ relative_coords[:, :, 1] += training_window_size[1] - 1
|
|
|
|
|
+ relative_coords[:, :, 0] *= 2 * training_window_size[1] - 1
|
|
|
|
|
+ relative_position_index = \
|
|
|
|
|
+ torch.zeros(size=(training_window_size[0] * training_window_size[1] + 1,) * 2,
|
|
|
|
|
+ dtype=relative_coords.dtype)
|
|
|
|
|
+ relative_position_index[1:, 1:] = relative_coords.sum(-1) # Wh*Ww, Wh*Ww
|
|
|
|
|
+ relative_position_index[0, 0:] = new_num_relative_distance - 3
|
|
|
|
|
+ relative_position_index[0:, 0] = new_num_relative_distance - 2
|
|
|
|
|
+ relative_position_index[0, 0] = new_num_relative_distance - 1
|
|
|
|
|
+
|
|
|
|
|
+ relative_position_bias = \
|
|
|
|
|
+ new_relative_position_bias_table[relative_position_index.view(-1)].view(
|
|
|
|
|
+ training_window_size[0] * training_window_size[1] + 1,
|
|
|
|
|
+ training_window_size[0] * training_window_size[1] + 1, -1) # Wh*Ww,Wh*Ww,nH
|
|
|
|
|
+ relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww
|
|
|
|
|
+ attn = attn + relative_position_bias.unsqueeze(0)
|
|
|
|
|
+
|
|
|
|
|
+ if rel_pos_bias is not None:
|
|
|
|
|
+ attn = attn + rel_pos_bias
|
|
|
|
|
+
|
|
|
|
|
+ attn = attn.softmax(dim=-1)
|
|
|
|
|
+ attn = self.attn_drop(attn)
|
|
|
|
|
+
|
|
|
|
|
+ x = (attn @ v).transpose(1, 2).reshape(B, N, -1)
|
|
|
|
|
+ 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., init_values=None, act_layer=nn.GELU, norm_layer=nn.LayerNorm,
|
|
|
|
|
+ window_size=None, attn_head_dim=None):
|
|
|
|
|
+ super().__init__()
|
|
|
|
|
+ self.norm1 = norm_layer(dim)
|
|
|
|
|
+ self.attn = Attention(
|
|
|
|
|
+ dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale,
|
|
|
|
|
+ attn_drop=attn_drop, proj_drop=drop, window_size=window_size, attn_head_dim=attn_head_dim)
|
|
|
|
|
+ # 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)
|
|
|
|
|
+
|
|
|
|
|
+ if init_values is not None:
|
|
|
|
|
+ self.gamma_1 = nn.Parameter(init_values * torch.ones((dim)), requires_grad=True)
|
|
|
|
|
+ self.gamma_2 = nn.Parameter(init_values * torch.ones((dim)), requires_grad=True)
|
|
|
|
|
+ else:
|
|
|
|
|
+ self.gamma_1, self.gamma_2 = None, None
|
|
|
|
|
+
|
|
|
|
|
+ def forward(self, x, rel_pos_bias=None, training_window_size=None):
|
|
|
|
|
+ if self.gamma_1 is None:
|
|
|
|
|
+ x = x + self.drop_path(
|
|
|
|
|
+ self.attn(self.norm1(x), rel_pos_bias=rel_pos_bias, training_window_size=training_window_size))
|
|
|
|
|
+ x = x + self.drop_path(self.mlp(self.norm2(x)))
|
|
|
|
|
+ else:
|
|
|
|
|
+ x = x + self.drop_path(self.gamma_1 * self.attn(self.norm1(x), rel_pos_bias=rel_pos_bias,
|
|
|
|
|
+ training_window_size=training_window_size))
|
|
|
|
|
+ x = x + self.drop_path(self.gamma_2 * self.mlp(self.norm2(x)))
|
|
|
|
|
+ return x
|
|
|
|
|
+
|
|
|
|
|
+
|
|
|
|
|
+class PatchEmbed(nn.Module):
|
|
|
|
|
+ """ Image to Patch Embedding
|
|
|
|
|
+ """
|
|
|
|
|
+
|
|
|
|
|
+ def __init__(self, img_size=[224, 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.patch_shape = (img_size[0] // patch_size[0], img_size[1] // patch_size[1])
|
|
|
|
|
+ self.num_patches_w = self.patch_shape[0]
|
|
|
|
|
+ self.num_patches_h = self.patch_shape[1]
|
|
|
|
|
+ # the so-called patch_shape is the patch shape during pre-training
|
|
|
|
|
+ self.img_size = img_size
|
|
|
|
|
+ self.patch_size = patch_size
|
|
|
|
|
+ self.num_patches = num_patches
|
|
|
|
|
+
|
|
|
|
|
+ self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size)
|
|
|
|
|
+
|
|
|
|
|
+ def forward(self, x, position_embedding=None, **kwargs):
|
|
|
|
|
+ # 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)
|
|
|
|
|
+ Hp, Wp = x.shape[2], x.shape[3]
|
|
|
|
|
+
|
|
|
|
|
+ if position_embedding is not None:
|
|
|
|
|
+ # interpolate the position embedding to the corresponding size
|
|
|
|
|
+ position_embedding = position_embedding.view(1, self.patch_shape[0], self.patch_shape[1], -1).permute(0, 3,
|
|
|
|
|
+ 1, 2)
|
|
|
|
|
+ position_embedding = F.interpolate(position_embedding, size=(Hp, Wp), mode='bicubic')
|
|
|
|
|
+ x = x + position_embedding
|
|
|
|
|
+
|
|
|
|
|
+ x = x.flatten(2).transpose(1, 2)
|
|
|
|
|
+ return x, (Hp, Wp)
|
|
|
|
|
+
|
|
|
|
|
+
|
|
|
|
|
+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, 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]))[-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.Linear(feature_dim, embed_dim)
|
|
|
|
|
+
|
|
|
|
|
+ def forward(self, x):
|
|
|
|
|
+ x = self.backbone(x)[-1]
|
|
|
|
|
+ x = x.flatten(2).transpose(1, 2)
|
|
|
|
|
+ x = self.proj(x)
|
|
|
|
|
+ return x
|
|
|
|
|
+
|
|
|
|
|
+
|
|
|
|
|
+class RelativePositionBias(nn.Module):
|
|
|
|
|
+
|
|
|
|
|
+ def __init__(self, window_size, num_heads):
|
|
|
|
|
+ super().__init__()
|
|
|
|
|
+ self.window_size = window_size
|
|
|
|
|
+ self.num_heads = num_heads
|
|
|
|
|
+ self.num_relative_distance = (2 * window_size[0] - 1) * (2 * window_size[1] - 1) + 3
|
|
|
|
|
+ self.relative_position_bias_table = nn.Parameter(
|
|
|
|
|
+ torch.zeros(self.num_relative_distance, num_heads)) # 2*Wh-1 * 2*Ww-1, nH
|
|
|
|
|
+ # cls to token & token 2 cls & cls to cls
|
|
|
|
|
+
|
|
|
|
|
+ # get pair-wise relative position index for each token inside the window
|
|
|
|
|
+ coords_h = torch.arange(window_size[0])
|
|
|
|
|
+ coords_w = torch.arange(window_size[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] += window_size[0] - 1 # shift to start from 0
|
|
|
|
|
+ relative_coords[:, :, 1] += window_size[1] - 1
|
|
|
|
|
+ relative_coords[:, :, 0] *= 2 * window_size[1] - 1
|
|
|
|
|
+ relative_position_index = \
|
|
|
|
|
+ torch.zeros(size=(window_size[0] * window_size[1] + 1,) * 2, dtype=relative_coords.dtype)
|
|
|
|
|
+ relative_position_index[1:, 1:] = relative_coords.sum(-1) # Wh*Ww, Wh*Ww
|
|
|
|
|
+ relative_position_index[0, 0:] = self.num_relative_distance - 3
|
|
|
|
|
+ relative_position_index[0:, 0] = self.num_relative_distance - 2
|
|
|
|
|
+ relative_position_index[0, 0] = self.num_relative_distance - 1
|
|
|
|
|
+
|
|
|
|
|
+ self.register_buffer("relative_position_index", relative_position_index)
|
|
|
|
|
+
|
|
|
|
|
+ # trunc_normal_(self.relative_position_bias_table, std=.02)
|
|
|
|
|
+
|
|
|
|
|
+ def forward(self, training_window_size):
|
|
|
|
|
+ if training_window_size == self.window_size:
|
|
|
|
|
+ relative_position_bias = \
|
|
|
|
|
+ self.relative_position_bias_table[self.relative_position_index.view(-1)].view(
|
|
|
|
|
+ self.window_size[0] * self.window_size[1] + 1,
|
|
|
|
|
+ self.window_size[0] * self.window_size[1] + 1, -1) # Wh*Ww,Wh*Ww,nH
|
|
|
|
|
+ relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww
|
|
|
|
|
+ else:
|
|
|
|
|
+ training_window_size = tuple(training_window_size.tolist())
|
|
|
|
|
+ new_num_relative_distance = (2 * training_window_size[0] - 1) * (2 * training_window_size[1] - 1) + 3
|
|
|
|
|
+ # new_num_relative_dis 为 所有可能的相对位置选项,包含cls-cls,tok-cls,与cls-tok
|
|
|
|
|
+ new_relative_position_bias_table = F.interpolate(
|
|
|
|
|
+ self.relative_position_bias_table[:-3, :].permute(1, 0).view(1, self.num_heads,
|
|
|
|
|
+ 2 * self.window_size[0] - 1,
|
|
|
|
|
+ 2 * self.window_size[1] - 1),
|
|
|
|
|
+ size=(2 * training_window_size[0] - 1, 2 * training_window_size[1] - 1), mode='bicubic',
|
|
|
|
|
+ align_corners=False)
|
|
|
|
|
+ new_relative_position_bias_table = new_relative_position_bias_table.view(self.num_heads,
|
|
|
|
|
+ new_num_relative_distance - 3).permute(
|
|
|
|
|
+ 1, 0)
|
|
|
|
|
+ new_relative_position_bias_table = torch.cat(
|
|
|
|
|
+ [new_relative_position_bias_table, self.relative_position_bias_table[-3::]], dim=0)
|
|
|
|
|
+
|
|
|
|
|
+ # get pair-wise relative position index for each token inside the window
|
|
|
|
|
+ coords_h = torch.arange(training_window_size[0])
|
|
|
|
|
+ coords_w = torch.arange(training_window_size[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] += training_window_size[0] - 1 # shift to start from 0
|
|
|
|
|
+ relative_coords[:, :, 1] += training_window_size[1] - 1
|
|
|
|
|
+ relative_coords[:, :, 0] *= 2 * training_window_size[1] - 1
|
|
|
|
|
+ relative_position_index = \
|
|
|
|
|
+ torch.zeros(size=(training_window_size[0] * training_window_size[1] + 1,) * 2,
|
|
|
|
|
+ dtype=relative_coords.dtype)
|
|
|
|
|
+ relative_position_index[1:, 1:] = relative_coords.sum(-1) # Wh*Ww, Wh*Ww
|
|
|
|
|
+ relative_position_index[0, 0:] = new_num_relative_distance - 3
|
|
|
|
|
+ relative_position_index[0:, 0] = new_num_relative_distance - 2
|
|
|
|
|
+ relative_position_index[0, 0] = new_num_relative_distance - 1
|
|
|
|
|
+
|
|
|
|
|
+ relative_position_bias = \
|
|
|
|
|
+ new_relative_position_bias_table[relative_position_index.view(-1)].view(
|
|
|
|
|
+ training_window_size[0] * training_window_size[1] + 1,
|
|
|
|
|
+ training_window_size[0] * training_window_size[1] + 1, -1) # Wh*Ww,Wh*Ww,nH
|
|
|
|
|
+ relative_position_bias = relative_position_bias.permute(2, 0, 1).contiguous() # nH, Wh*Ww, Wh*Ww
|
|
|
|
|
+
|
|
|
|
|
+ return relative_position_bias
|
|
|
|
|
+
|
|
|
|
|
+
|
|
|
|
|
+class BEiT(nn.Module):
|
|
|
|
|
+ """ Vision Transformer with support for patch or hybrid CNN input stage
|
|
|
|
|
+ """
|
|
|
|
|
+
|
|
|
|
|
+ def __init__(self,
|
|
|
|
|
+ img_size=[224, 224],
|
|
|
|
|
+ patch_size=16,
|
|
|
|
|
+ in_chans=3,
|
|
|
|
|
+ num_classes=80,
|
|
|
|
|
+ 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.,
|
|
|
|
|
+ hybrid_backbone=None,
|
|
|
|
|
+ norm_layer=None,
|
|
|
|
|
+ init_values=None,
|
|
|
|
|
+ use_abs_pos_emb=False,
|
|
|
|
|
+ use_rel_pos_bias=False,
|
|
|
|
|
+ use_shared_rel_pos_bias=False,
|
|
|
|
|
+ use_checkpoint=True,
|
|
|
|
|
+ pretrained=None,
|
|
|
|
|
+ out_features=None,
|
|
|
|
|
+ ):
|
|
|
|
|
+
|
|
|
|
|
+ super(BEiT, self).__init__()
|
|
|
|
|
+
|
|
|
|
|
+ norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6)
|
|
|
|
|
+ self.num_classes = num_classes
|
|
|
|
|
+ self.num_features = self.embed_dim = embed_dim # num_features for consistency with other models
|
|
|
|
|
+ self.use_checkpoint = use_checkpoint
|
|
|
|
|
+
|
|
|
|
|
+ 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.out_features = out_features
|
|
|
|
|
+ self.out_indices = [int(name[5:]) for name in out_features]
|
|
|
|
|
+
|
|
|
|
|
+ self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
|
|
|
|
|
+ # self.mask_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
|
|
|
|
|
+ if use_abs_pos_emb:
|
|
|
|
|
+ self.pos_embed = nn.Parameter(torch.zeros(1, num_patches + 1, embed_dim))
|
|
|
|
|
+ else:
|
|
|
|
|
+ self.pos_embed = None
|
|
|
|
|
+ self.pos_drop = nn.Dropout(p=drop_rate)
|
|
|
|
|
+
|
|
|
|
|
+ self.use_shared_rel_pos_bias = use_shared_rel_pos_bias
|
|
|
|
|
+ if use_shared_rel_pos_bias:
|
|
|
|
|
+ self.rel_pos_bias = RelativePositionBias(window_size=self.patch_embed.patch_shape, num_heads=num_heads)
|
|
|
|
|
+ else:
|
|
|
|
|
+ self.rel_pos_bias = None
|
|
|
|
|
+
|
|
|
|
|
+ dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)] # stochastic depth decay rule
|
|
|
|
|
+ self.use_rel_pos_bias = use_rel_pos_bias
|
|
|
|
|
+ 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,
|
|
|
|
|
+ init_values=init_values, window_size=self.patch_embed.patch_shape if use_rel_pos_bias else None)
|
|
|
|
|
+ for i in range(depth)])
|
|
|
|
|
+
|
|
|
|
|
+ # trunc_normal_(self.mask_token, std=.02)
|
|
|
|
|
+
|
|
|
|
|
+ if patch_size == 16:
|
|
|
|
|
+ self.fpn1 = nn.Sequential(
|
|
|
|
|
+ nn.ConvTranspose2d(embed_dim, embed_dim, kernel_size=2, stride=2),
|
|
|
|
|
+ # nn.SyncBatchNorm(embed_dim),
|
|
|
|
|
+ nn.BatchNorm2d(embed_dim),
|
|
|
|
|
+ nn.GELU(),
|
|
|
|
|
+ nn.ConvTranspose2d(embed_dim, embed_dim, kernel_size=2, stride=2),
|
|
|
|
|
+ )
|
|
|
|
|
+
|
|
|
|
|
+ self.fpn2 = nn.Sequential(
|
|
|
|
|
+ nn.ConvTranspose2d(embed_dim, embed_dim, kernel_size=2, stride=2),
|
|
|
|
|
+ )
|
|
|
|
|
+
|
|
|
|
|
+ self.fpn3 = nn.Identity()
|
|
|
|
|
+
|
|
|
|
|
+ self.fpn4 = nn.MaxPool2d(kernel_size=2, stride=2)
|
|
|
|
|
+ elif patch_size == 8:
|
|
|
|
|
+ self.fpn1 = nn.Sequential(
|
|
|
|
|
+ nn.ConvTranspose2d(embed_dim, embed_dim, kernel_size=2, stride=2),
|
|
|
|
|
+ )
|
|
|
|
|
+
|
|
|
|
|
+ self.fpn2 = nn.Identity()
|
|
|
|
|
+
|
|
|
|
|
+ self.fpn3 = nn.Sequential(
|
|
|
|
|
+ nn.MaxPool2d(kernel_size=2, stride=2),
|
|
|
|
|
+ )
|
|
|
|
|
+
|
|
|
|
|
+ self.fpn4 = nn.Sequential(
|
|
|
|
|
+ nn.MaxPool2d(kernel_size=4, stride=4),
|
|
|
|
|
+ )
|
|
|
|
|
+
|
|
|
|
|
+ if self.pos_embed is not None:
|
|
|
|
|
+ trunc_normal_(self.pos_embed, std=.02)
|
|
|
|
|
+ trunc_normal_(self.cls_token, std=.02)
|
|
|
|
|
+ self.apply(self._init_weights)
|
|
|
|
|
+ self.fix_init_weight()
|
|
|
|
|
+
|
|
|
|
|
+ def fix_init_weight(self):
|
|
|
|
|
+ def rescale(param, layer_id):
|
|
|
|
|
+ param.div_(math.sqrt(2.0 * layer_id))
|
|
|
|
|
+
|
|
|
|
|
+ for layer_id, layer in enumerate(self.blocks):
|
|
|
|
|
+ rescale(layer.attn.proj.weight.data, layer_id + 1)
|
|
|
|
|
+ rescale(layer.mlp.fc2.weight.data, layer_id + 1)
|
|
|
|
|
+
|
|
|
|
|
+ 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)
|
|
|
|
|
+
|
|
|
|
|
+ '''
|
|
|
|
|
+ def init_weights(self):
|
|
|
|
|
+ """Initialize the weights in backbone.
|
|
|
|
|
+
|
|
|
|
|
+ Args:
|
|
|
|
|
+ pretrained (str, optional): Path to pre-trained weights.
|
|
|
|
|
+ Defaults to None.
|
|
|
|
|
+ """
|
|
|
|
|
+ logger = get_root_logger()
|
|
|
|
|
+
|
|
|
|
|
+ if self.pos_embed is not None:
|
|
|
|
|
+ trunc_normal_(self.pos_embed, std=.02)
|
|
|
|
|
+ trunc_normal_(self.cls_token, std=.02)
|
|
|
|
|
+ self.apply(self._init_weights)
|
|
|
|
|
+ self.fix_init_weight()
|
|
|
|
|
+
|
|
|
|
|
+ if self.init_cfg is None:
|
|
|
|
|
+ logger.warn(f'No pre-trained weights for '
|
|
|
|
|
+ f'{self.__class__.__name__}, '
|
|
|
|
|
+ f'training start from scratch')
|
|
|
|
|
+ else:
|
|
|
|
|
+ assert 'checkpoint' in self.init_cfg, f'Only support ' \
|
|
|
|
|
+ f'specify `Pretrained` in ' \
|
|
|
|
|
+ f'`init_cfg` in ' \
|
|
|
|
|
+ f'{self.__class__.__name__} '
|
|
|
|
|
+ logger.info(f"Will load ckpt from {self.init_cfg['checkpoint']}")
|
|
|
|
|
+ load_checkpoint(self,
|
|
|
|
|
+ filename=self.init_cfg['checkpoint'],
|
|
|
|
|
+ strict=False,
|
|
|
|
|
+ logger=logger,
|
|
|
|
|
+ beit_spec_expand_rel_pos = self.use_rel_pos_bias,
|
|
|
|
|
+ )
|
|
|
|
|
+ '''
|
|
|
|
|
+
|
|
|
|
|
+ def get_num_layers(self):
|
|
|
|
|
+ return len(self.blocks)
|
|
|
|
|
+
|
|
|
|
|
+ @torch.jit.ignore
|
|
|
|
|
+ def no_weight_decay(self):
|
|
|
|
|
+ return {'pos_embed', 'cls_token'}
|
|
|
|
|
+
|
|
|
|
|
+ def forward_features(self, x):
|
|
|
|
|
+ B, C, H, W = x.shape
|
|
|
|
|
+ x, (Hp, Wp) = self.patch_embed(x, self.pos_embed[:, 1:, :] if self.pos_embed is not None else None)
|
|
|
|
|
+ # Hp, Wp are HW for patches
|
|
|
|
|
+ batch_size, seq_len, _ = x.size()
|
|
|
|
|
+
|
|
|
|
|
+ cls_tokens = self.cls_token.expand(batch_size, -1, -1) # stole cls_tokens impl from Phil Wang, thanks
|
|
|
|
|
+ if self.pos_embed is not None:
|
|
|
|
|
+ cls_tokens = cls_tokens + self.pos_embed[:, :1, :]
|
|
|
|
|
+ x = torch.cat((cls_tokens, x), dim=1)
|
|
|
|
|
+ x = self.pos_drop(x)
|
|
|
|
|
+
|
|
|
|
|
+ features = []
|
|
|
|
|
+ training_window_size = torch.tensor([Hp, Wp])
|
|
|
|
|
+
|
|
|
|
|
+ rel_pos_bias = self.rel_pos_bias(training_window_size) if self.rel_pos_bias is not None else None
|
|
|
|
|
+
|
|
|
|
|
+ for i, blk in enumerate(self.blocks):
|
|
|
|
|
+ if self.use_checkpoint:
|
|
|
|
|
+ x = checkpoint.checkpoint(blk, x, rel_pos_bias, training_window_size)
|
|
|
|
|
+ else:
|
|
|
|
|
+ x = blk(x, rel_pos_bias=rel_pos_bias, training_window_size=training_window_size)
|
|
|
|
|
+ if i in self.out_indices:
|
|
|
|
|
+ xp = x[:, 1:, :].permute(0, 2, 1).reshape(B, -1, Hp, Wp)
|
|
|
|
|
+ features.append(xp.contiguous())
|
|
|
|
|
+
|
|
|
|
|
+ ops = [self.fpn1, self.fpn2, self.fpn3, self.fpn4]
|
|
|
|
|
+ for i in range(len(features)):
|
|
|
|
|
+ features[i] = ops[i](features[i])
|
|
|
|
|
+
|
|
|
|
|
+ feat_out = {}
|
|
|
|
|
+
|
|
|
|
|
+ for name, value in zip(self.out_features, features):
|
|
|
|
|
+ feat_out[name] = value
|
|
|
|
|
+
|
|
|
|
|
+ return feat_out
|
|
|
|
|
+
|
|
|
|
|
+ def forward(self, x):
|
|
|
|
|
+ x = self.forward_features(x)
|
|
|
|
|
+ return x
|
|
|
|
|
+
|
|
|
|
|
+
|
|
|
|
|
+def beit_base_patch16(pretrained=False, **kwargs):
|
|
|
|
|
+ model = BEiT(
|
|
|
|
|
+ patch_size=16,
|
|
|
|
|
+ embed_dim=768,
|
|
|
|
|
+ depth=12,
|
|
|
|
|
+ num_heads=12,
|
|
|
|
|
+ mlp_ratio=4,
|
|
|
|
|
+ qkv_bias=True,
|
|
|
|
|
+ norm_layer=partial(nn.LayerNorm, eps=1e-6),
|
|
|
|
|
+ init_values=None,
|
|
|
|
|
+ **kwargs)
|
|
|
|
|
+ model.default_cfg = _cfg()
|
|
|
|
|
+ return model
|
|
|
|
|
+
|
|
|
|
|
+def beit_large_patch16(pretrained=False, **kwargs):
|
|
|
|
|
+ model = BEiT(
|
|
|
|
|
+ patch_size=16,
|
|
|
|
|
+ embed_dim=1024,
|
|
|
|
|
+ depth=24,
|
|
|
|
|
+ num_heads=16,
|
|
|
|
|
+ mlp_ratio=4,
|
|
|
|
|
+ qkv_bias=True,
|
|
|
|
|
+ norm_layer=partial(nn.LayerNorm, eps=1e-6),
|
|
|
|
|
+ init_values=None,
|
|
|
|
|
+ **kwargs)
|
|
|
|
|
+ model.default_cfg = _cfg()
|
|
|
|
|
+ return model
|
|
|
|
|
+
|
|
|
|
|
+def dit_base_patch16(pretrained=False, **kwargs):
|
|
|
|
|
+ model = BEiT(
|
|
|
|
|
+ patch_size=16,
|
|
|
|
|
+ embed_dim=768,
|
|
|
|
|
+ depth=12,
|
|
|
|
|
+ num_heads=12,
|
|
|
|
|
+ mlp_ratio=4,
|
|
|
|
|
+ qkv_bias=True,
|
|
|
|
|
+ norm_layer=partial(nn.LayerNorm, eps=1e-6),
|
|
|
|
|
+ init_values=0.1,
|
|
|
|
|
+ **kwargs)
|
|
|
|
|
+ model.default_cfg = _cfg()
|
|
|
|
|
+ return model
|
|
|
|
|
+
|
|
|
|
|
+def dit_large_patch16(pretrained=False, **kwargs):
|
|
|
|
|
+ model = BEiT(
|
|
|
|
|
+ patch_size=16,
|
|
|
|
|
+ embed_dim=1024,
|
|
|
|
|
+ depth=24,
|
|
|
|
|
+ num_heads=16,
|
|
|
|
|
+ mlp_ratio=4,
|
|
|
|
|
+ qkv_bias=True,
|
|
|
|
|
+ norm_layer=partial(nn.LayerNorm, eps=1e-6),
|
|
|
|
|
+ init_values=1e-5,
|
|
|
|
|
+ **kwargs)
|
|
|
|
|
+ model.default_cfg = _cfg()
|
|
|
|
|
+ return model
|
|
|
|
|
+
|
|
|
|
|
+if __name__ == '__main__':
|
|
|
|
|
+ model = BEiT(use_checkpoint=True, use_shared_rel_pos_bias=True)
|
|
|
|
|
+ model = model.to("cuda:0")
|
|
|
|
|
+ input1 = torch.rand(2, 3, 512, 762).to("cuda:0")
|
|
|
|
|
+ input2 = torch.rand(2, 3, 800, 1200).to("cuda:0")
|
|
|
|
|
+ input3 = torch.rand(2, 3, 720, 1000).to("cuda:0")
|
|
|
|
|
+ output1 = model(input1)
|
|
|
|
|
+ output2 = model(input2)
|
|
|
|
|
+ output3 = model(input3)
|
|
|
|
|
+ print("all done")
|