From 105e97f24070ed0a414da122153b1913974f0255 Mon Sep 17 00:00:00 2001 From: Phil Wang Date: Mon, 5 Dec 2022 10:47:02 -0800 Subject: [PATCH] need simple vit with patch dropout for another project --- setup.py | 2 +- vit_pytorch/simple_vit.py | 21 +++ vit_pytorch/simple_vit_with_patch_dropout.py | 141 +++++++++++++++++++ 3 files changed, 163 insertions(+), 1 deletion(-) create mode 100644 vit_pytorch/simple_vit_with_patch_dropout.py diff --git a/setup.py b/setup.py index c05890c..ab1e0ae 100644 --- a/setup.py +++ b/setup.py @@ -3,7 +3,7 @@ setup( name = 'vit-pytorch', packages = find_packages(exclude=['examples']), - version = '0.40.1', + version = '0.40.2', license='MIT', description = 'Vision Transformer (ViT) - Pytorch', long_description_content_type = 'text/markdown', diff --git a/vit_pytorch/simple_vit.py b/vit_pytorch/simple_vit.py index 2db187e..5bd6c49 100644 --- a/vit_pytorch/simple_vit.py +++ b/vit_pytorch/simple_vit.py @@ -22,6 +22,27 @@ def posemb_sincos_2d(patches, temperature = 10000, dtype = torch.float32): pe = torch.cat((x.sin(), x.cos(), y.sin(), y.cos()), dim = 1) return pe.type(dtype) +# patch dropout + +class PatchDropout(nn.Module): + def __init__(self, prob): + super().__init__() + assert 0 <= prob < 1. + self.prob = prob + + def forward(self, x): + if not self.training or self.prob == 0.: + return x + + b, n, _, device = *x.shape, x.device + + batch_indices = torch.arange(b, device = device) + batch_indices = rearrange(batch_indices, '... -> ... 1') + num_patches_keep = max(1, int(n * (1 - self.prob))) + patch_indices_keep = torch.randn(b, n, device = device).topk(num_patches_keep, dim = -1).indices + + return x[batch_indices, patch_indices_keep] + # classes class FeedForward(nn.Module): diff --git a/vit_pytorch/simple_vit_with_patch_dropout.py b/vit_pytorch/simple_vit_with_patch_dropout.py new file mode 100644 index 0000000..87104ef --- /dev/null +++ b/vit_pytorch/simple_vit_with_patch_dropout.py @@ -0,0 +1,141 @@ +import torch +from torch import nn + +from einops import rearrange +from einops.layers.torch import Rearrange + +# helpers + +def pair(t): + return t if isinstance(t, tuple) else (t, t) + +def posemb_sincos_2d(patches, temperature = 10000, dtype = torch.float32): + _, h, w, dim, device, dtype = *patches.shape, patches.device, patches.dtype + + y, x = torch.meshgrid(torch.arange(h, device = device), torch.arange(w, device = device), indexing = 'ij') + assert (dim % 4) == 0, 'feature dimension must be multiple of 4 for sincos emb' + omega = torch.arange(dim // 4, device = device) / (dim // 4 - 1) + omega = 1. / (temperature ** omega) + + y = y.flatten()[:, None] * omega[None, :] + x = x.flatten()[:, None] * omega[None, :] + pe = torch.cat((x.sin(), x.cos(), y.sin(), y.cos()), dim = 1) + return pe.type(dtype) + +# patch dropout + +class PatchDropout(nn.Module): + def __init__(self, prob): + super().__init__() + assert 0 <= prob < 1. + self.prob = prob + + def forward(self, x): + if not self.training or self.prob == 0.: + return x + + b, n, _, device = *x.shape, x.device + + batch_indices = torch.arange(b, device = device) + batch_indices = rearrange(batch_indices, '... -> ... 1') + num_patches_keep = max(1, int(n * (1 - self.prob))) + patch_indices_keep = torch.randn(b, n, device = device).topk(num_patches_keep, dim = -1).indices + + return x[batch_indices, patch_indices_keep] + +# classes + +class FeedForward(nn.Module): + def __init__(self, dim, hidden_dim): + super().__init__() + self.net = nn.Sequential( + nn.LayerNorm(dim), + nn.Linear(dim, hidden_dim), + nn.GELU(), + nn.Linear(hidden_dim, dim), + ) + def forward(self, x): + return self.net(x) + +class Attention(nn.Module): + def __init__(self, dim, heads = 8, dim_head = 64): + super().__init__() + inner_dim = dim_head * heads + self.heads = heads + self.scale = dim_head ** -0.5 + self.norm = nn.LayerNorm(dim) + + self.attend = nn.Softmax(dim = -1) + + self.to_qkv = nn.Linear(dim, inner_dim * 3, bias = False) + self.to_out = nn.Linear(inner_dim, dim, bias = False) + + def forward(self, x): + x = self.norm(x) + + qkv = self.to_qkv(x).chunk(3, dim = -1) + q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> b h n d', h = self.heads), qkv) + + dots = torch.matmul(q, k.transpose(-1, -2)) * self.scale + + attn = self.attend(dots) + + out = torch.matmul(attn, v) + out = rearrange(out, 'b h n d -> b n (h d)') + return self.to_out(out) + +class Transformer(nn.Module): + def __init__(self, dim, depth, heads, dim_head, mlp_dim): + super().__init__() + self.layers = nn.ModuleList([]) + for _ in range(depth): + self.layers.append(nn.ModuleList([ + Attention(dim, heads = heads, dim_head = dim_head), + FeedForward(dim, mlp_dim) + ])) + def forward(self, x): + for attn, ff in self.layers: + x = attn(x) + x + x = ff(x) + x + return x + +class SimpleViT(nn.Module): + def __init__(self, *, image_size, patch_size, num_classes, dim, depth, heads, mlp_dim, channels = 3, dim_head = 64, patch_dropout = 0.5): + super().__init__() + image_height, image_width = pair(image_size) + patch_height, patch_width = pair(patch_size) + + assert image_height % patch_height == 0 and image_width % patch_width == 0, 'Image dimensions must be divisible by the patch size.' + + num_patches = (image_height // patch_height) * (image_width // patch_width) + patch_dim = channels * patch_height * patch_width + + self.to_patch_embedding = nn.Sequential( + Rearrange('b c (h p1) (w p2) -> b h w (p1 p2 c)', p1 = patch_height, p2 = patch_width), + nn.Linear(patch_dim, dim), + ) + + self.patch_dropout = PatchDropout(patch_dropout) + + self.transformer = Transformer(dim, depth, heads, dim_head, mlp_dim) + + self.to_latent = nn.Identity() + self.linear_head = nn.Sequential( + nn.LayerNorm(dim), + nn.Linear(dim, num_classes) + ) + + def forward(self, img): + *_, h, w, dtype = *img.shape, img.dtype + + x = self.to_patch_embedding(img) + pe = posemb_sincos_2d(x) + x = rearrange(x, 'b ... d -> b (...) d') + pe + + x = self.patch_dropout(x) + + x = self.transformer(x) + x = x.mean(dim = 1) + + x = self.to_latent(x) + return self.linear_head(x)