diff --git a/timm/models/convit.py b/timm/models/convit.py index 31c05df35a03383f6e68f595fc47a0538515c47a..f6ae3ec108392f406ac5a72e44565cea02ef9114 100644 --- a/timm/models/convit.py +++ b/timm/models/convit.py @@ -1,6 +1,24 @@ -"""These modules are adapted from those of timm, see -https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py +""" ConViT Model + +@article{d2021convit, + title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases}, + author={d'Ascoli, St{\'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent}, + journal={arXiv preprint arXiv:2103.10697}, + year={2021} +} + +Paper link: https://arxiv.org/abs/2103.10697 +Original code: https://github.com/facebookresearch/convit, original copyright below """ +# Copyright (c) 2015-present, Facebook, Inc. +# All rights reserved. +# +# This source code is licensed under the CC-by-NC license found in the +# LICENSE file in the root directory of this source tree. +# +'''These modules are adapted from those of timm, see +https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py +''' import torch import torch.nn as nn @@ -9,8 +27,9 @@ import torch.nn.functional as F from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD from .helpers import build_model_with_cfg -from timm.models.layers import DropPath, to_2tuple, trunc_normal_ -from timm.models.registry import register_model +from .layers import DropPath, to_2tuple, trunc_normal_, PatchEmbed, Mlp +from .registry import register_model +from .vision_transformer_hybrid import HybridEmbed import torch import torch.nn as nn @@ -29,7 +48,7 @@ def _cfg(url='', **kwargs): default_cfgs = { # ConViT 'convit_tiny': _cfg( - url="https://dl.fbaipublicfiles.com/convit/convit_tiny.pth"), + url="https://dl.fbaipublicfiles.com/convit/convit_tiny.pth"), 'convit_small': _cfg( url="https://dl.fbaipublicfiles.com/convit/convit_small.pth"), 'convit_base': _cfg( @@ -37,71 +56,31 @@ default_cfgs = { } -class Mlp(nn.Module): - def __init__(self, in_features, hidden_features=None, out_features=None, act_layer=nn.GELU, drop=0.): - super().__init__() - out_features = out_features or in_features - hidden_features = hidden_features or in_features - self.fc1 = nn.Linear(in_features, hidden_features) - self.act = act_layer() - self.fc2 = nn.Linear(hidden_features, out_features) - self.drop = nn.Dropout(drop) - self.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) - - 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 GPSA(nn.Module): def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0., - locality_strength=1., use_local_init=True): + locality_strength=1.): super().__init__() self.num_heads = num_heads self.dim = dim head_dim = dim // num_heads self.scale = qk_scale or head_dim ** -0.5 + self.locality_strength = locality_strength + + self.qk = nn.Linear(dim, dim * 2, bias=qkv_bias) + self.v = nn.Linear(dim, dim, bias=qkv_bias) - self.qk = nn.Linear(dim, dim * 2, bias=qkv_bias) - self.v = nn.Linear(dim, dim, bias=qkv_bias) - self.attn_drop = nn.Dropout(attn_drop) self.proj = nn.Linear(dim, dim) self.pos_proj = nn.Linear(3, num_heads) self.proj_drop = nn.Dropout(proj_drop) self.locality_strength = locality_strength self.gating_param = nn.Parameter(torch.ones(self.num_heads)) - self.apply(self._init_weights) - if use_local_init: - self.local_init(locality_strength=locality_strength) + self.rel_indices: torch.Tensor = torch.zeros(1, 1, 1, 3) # silly torchscript hack, won't work with None - 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 forward(self, x): B, N, C = x.shape - if not hasattr(self, 'rel_indices') or self.rel_indices.size(1)!=N: - self.get_rel_indices(N) - + if self.rel_indices is None or self.rel_indices.shape[1] != N: + self.rel_indices = self.get_rel_indices(N) attn = self.get_attention(x) v = self.v(x).reshape(B, N, self.num_heads, C // self.num_heads).permute(0, 2, 1, 3) x = (attn @ v).transpose(1, 2).reshape(B, N, C) @@ -110,61 +89,58 @@ class GPSA(nn.Module): return x def get_attention(self, x): - B, N, C = x.shape + B, N, C = x.shape qk = self.qk(x).reshape(B, N, 2, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4) q, k = qk[0], qk[1] - pos_score = self.rel_indices.expand(B, -1, -1,-1) - pos_score = self.pos_proj(pos_score).permute(0,3,1,2) + pos_score = self.rel_indices.expand(B, -1, -1, -1) + pos_score = self.pos_proj(pos_score).permute(0, 3, 1, 2) patch_score = (q @ k.transpose(-2, -1)) * self.scale patch_score = patch_score.softmax(dim=-1) pos_score = pos_score.softmax(dim=-1) - gating = self.gating_param.view(1,-1,1,1) - attn = (1.-torch.sigmoid(gating)) * patch_score + torch.sigmoid(gating) * pos_score + gating = self.gating_param.view(1, -1, 1, 1) + attn = (1. - torch.sigmoid(gating)) * patch_score + torch.sigmoid(gating) * pos_score attn /= attn.sum(dim=-1).unsqueeze(-1) attn = self.attn_drop(attn) return attn - def get_attention_map(self, x, return_map = False): - - attn_map = self.get_attention(x).mean(0) # average over batch - distances = self.rel_indices.squeeze()[:,:,-1]**.5 - dist = torch.einsum('nm,hnm->h', (distances, attn_map)) - dist /= distances.size(0) + def get_attention_map(self, x, return_map=False): + attn_map = self.get_attention(x).mean(0) # average over batch + distances = self.rel_indices.squeeze()[:, :, -1] ** .5 + dist = torch.einsum('nm,hnm->h', (distances, attn_map)) / distances.size(0) if return_map: return dist, attn_map else: return dist - - def local_init(self, locality_strength=1.): - + + def local_init(self): self.v.weight.data.copy_(torch.eye(self.dim)) - locality_distance = 1 #max(1,1/locality_strength**.5) - - kernel_size = int(self.num_heads**.5) - center = (kernel_size-1)/2 if kernel_size%2==0 else kernel_size//2 + locality_distance = 1 # max(1,1/locality_strength**.5) + + kernel_size = int(self.num_heads ** .5) + center = (kernel_size - 1) / 2 if kernel_size % 2 == 0 else kernel_size // 2 for h1 in range(kernel_size): for h2 in range(kernel_size): - position = h1+kernel_size*h2 - self.pos_proj.weight.data[position,2] = -1 - self.pos_proj.weight.data[position,1] = 2*(h1-center)*locality_distance - self.pos_proj.weight.data[position,0] = 2*(h2-center)*locality_distance - self.pos_proj.weight.data *= locality_strength - - def get_rel_indices(self, num_patches): - img_size = int(num_patches**.5) - rel_indices = torch.zeros(1, num_patches, num_patches, 3) - ind = torch.arange(img_size).view(1,-1) - torch.arange(img_size).view(-1, 1) - indx = ind.repeat(img_size,img_size) - indy = ind.repeat_interleave(img_size,dim=0).repeat_interleave(img_size,dim=1) - indd = indx**2 + indy**2 - rel_indices[:,:,:,2] = indd.unsqueeze(0) - rel_indices[:,:,:,1] = indy.unsqueeze(0) - rel_indices[:,:,:,0] = indx.unsqueeze(0) + position = h1 + kernel_size * h2 + self.pos_proj.weight.data[position, 2] = -1 + self.pos_proj.weight.data[position, 1] = 2 * (h1 - center) * locality_distance + self.pos_proj.weight.data[position, 0] = 2 * (h2 - center) * locality_distance + self.pos_proj.weight.data *= self.locality_strength + + def get_rel_indices(self, num_patches: int) -> torch.Tensor: + img_size = int(num_patches ** .5) + rel_indices = torch.zeros(1, num_patches, num_patches, 3) + ind = torch.arange(img_size).view(1, -1) - torch.arange(img_size).view(-1, 1) + indx = ind.repeat(img_size, img_size) + indy = ind.repeat_interleave(img_size, dim=0).repeat_interleave(img_size, dim=1) + indd = indx ** 2 + indy ** 2 + rel_indices[:, :, :, 2] = indd.unsqueeze(0) + rel_indices[:, :, :, 1] = indy.unsqueeze(0) + rel_indices[:, :, :, 0] = indx.unsqueeze(0) device = self.qk.weight.device - self.rel_indices = rel_indices.to(device) + return rel_indices.to(device) + - class MHSA(nn.Module): def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0.): super().__init__() @@ -176,41 +152,28 @@ class MHSA(nn.Module): self.attn_drop = nn.Dropout(attn_drop) self.proj = nn.Linear(dim, dim) self.proj_drop = nn.Dropout(proj_drop) - 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) - def get_attention_map(self, x, return_map = False): + def get_attention_map(self, x, return_map=False): 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] attn_map = (q @ k.transpose(-2, -1)) * self.scale attn_map = attn_map.softmax(dim=-1).mean(0) - img_size = int(N**.5) - ind = torch.arange(img_size).view(1,-1) - torch.arange(img_size).view(-1, 1) - indx = ind.repeat(img_size,img_size) - indy = ind.repeat_interleave(img_size,dim=0).repeat_interleave(img_size,dim=1) - indd = indx**2 + indy**2 - distances = indd**.5 + img_size = int(N ** .5) + ind = torch.arange(img_size).view(1, -1) - torch.arange(img_size).view(-1, 1) + indx = ind.repeat(img_size, img_size) + indy = ind.repeat_interleave(img_size, dim=0).repeat_interleave(img_size, dim=1) + indd = indx ** 2 + indy ** 2 + distances = indd ** .5 distances = distances.to('cuda') - dist = torch.einsum('nm,hnm->h', (distances, attn_map)) - dist /= N - + dist = torch.einsum('nm,hnm->h', (distances, attn_map)) / N if return_map: return dist, attn_map else: return dist - def forward(self, x): 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) @@ -228,15 +191,19 @@ class MHSA(nn.Module): class Block(nn.Module): - def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None, drop=0., attn_drop=0., + 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, use_gpsa=True, **kwargs): super().__init__() self.norm1 = norm_layer(dim) self.use_gpsa = use_gpsa if self.use_gpsa: - self.attn = GPSA(dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop, **kwargs) + self.attn = GPSA( + dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, + proj_drop=drop, **kwargs) else: - self.attn = MHSA(dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=drop, **kwargs) + self.attn = MHSA( + dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, + proj_drop=drop, **kwargs) 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) @@ -246,75 +213,12 @@ class Block(nn.Module): x = x + self.drop_path(self.attn(self.norm1(x))) x = x + self.drop_path(self.mlp(self.norm2(x))) return x - - -class PatchEmbed(nn.Module): - """ Image to Patch Embedding, from timm - """ - 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.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=patch_size) - self.apply(self._init_weights) - - 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).flatten(2).transpose(1, 2) - return x - - 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) - - -class HybridEmbed(nn.Module): - """ CNN Feature Map Embedding, from timm - """ - 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.Linear(feature_dim, embed_dim) - self.apply(self._init_weights) - - def forward(self, x): - x = self.backbone(x)[-1] - x = x.flatten(2).transpose(1, 2) - x = self.proj(x) - return x class ConViT(nn.Module): """ Vision Transformer with support for patch or hybrid CNN input stage """ + 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., hybrid_backbone=None, norm_layer=nn.LayerNorm, global_pool=None, @@ -335,7 +239,7 @@ class ConViT(nn.Module): img_size=img_size, patch_size=patch_size, in_chans=in_chans, embed_dim=embed_dim) num_patches = self.patch_embed.num_patches self.num_patches = num_patches - + self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim)) self.pos_drop = nn.Dropout(p=drop_rate) @@ -350,7 +254,7 @@ class ConViT(nn.Module): drop=drop_rate, attn_drop=attn_drop_rate, drop_path=dpr[i], norm_layer=norm_layer, use_gpsa=True, locality_strength=locality_strength) - if i 0 else nn.Identity() trunc_normal_(self.cls_token, std=.02) - self.head.apply(self._init_weights) + self.apply(self._init_weights) + for n, m in self.named_modules(): + if hasattr(m, 'local_init'): + m.local_init() def _init_weights(self, m): if isinstance(m, nn.Linear): @@ -395,8 +302,8 @@ class ConViT(nn.Module): x = x + self.pos_embed x = self.pos_drop(x) - for u,blk in enumerate(self.blocks): - if u == self.local_up_to_layer : + for u, blk in enumerate(self.blocks): + if u == self.local_up_to_layer: x = torch.cat((cls_tokens, x), dim=1) x = blk(x) @@ -415,30 +322,29 @@ def _create_convit(variant, pretrained=False, **kwargs): default_cfg=default_cfgs[variant], **kwargs) - + @register_model def convit_tiny(pretrained=False, **kwargs): model_args = dict( - local_up_to_layer=10, locality_strength=1.0, embed_dim=48, + local_up_to_layer=10, locality_strength=1.0, embed_dim=48, num_heads=4, norm_layer=partial(nn.LayerNorm, eps=1e-6), **kwargs) - model = _create_convit( - variant='convit_tiny', pretrained=pretrained, **model_args) + model = _create_convit(variant='convit_tiny', pretrained=pretrained, **model_args) return model + @register_model def convit_small(pretrained=False, **kwargs): model_args = dict( - local_up_to_layer=10, locality_strength=1.0, embed_dim=48, + local_up_to_layer=10, locality_strength=1.0, embed_dim=48, num_heads=9, norm_layer=partial(nn.LayerNorm, eps=1e-6), **kwargs) - model = _create_convit( - variant='convit_small', pretrained=pretrained, **model_args) + model = _create_convit(variant='convit_small', pretrained=pretrained, **model_args) return model + @register_model def convit_base(pretrained=False, **kwargs): model_args = dict( - local_up_to_layer=10, locality_strength=1.0, embed_dim=48, + local_up_to_layer=10, locality_strength=1.0, embed_dim=48, num_heads=16, norm_layer=partial(nn.LayerNorm, eps=1e-6), **kwargs) - model = _create_convit( - variant='convit_base', pretrained=pretrained, **model_args) + model = _create_convit(variant='convit_base', pretrained=pretrained, **model_args) return model