File size: 7,218 Bytes
d39a8f8 8c692c3 d39a8f8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 |
from transformers import PreTrainedModel
from .configuration_vitmix import ViTMixConfig
# Model architecture gracefully stolen from lucidrains https://github.com/lucidrains/vit-pytorch/blob/main/vit_pytorch/simple_vit.py
import torch
from torch import nn
from einops import rearrange
from einops.layers.torch import Rearrange
from st_moe_pytorch import SparseMoEBlock, MoE
# I thin this is 'including a copy of this notice'... tell me if I'm wrong
####################################
#MIT License
#Copyright (c) 2020 Phil Wang
#Permission is hereby granted, free of charge, to any person obtaining a copy
#of this software and associated documentation files (the "Software"), to deal
#in the Software without restriction, including without limitation the rights
#to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
#copies of the Software, and to permit persons to whom the Software is
#furnished to do so, subject to the following conditions:
#The above copyright notice and this permission notice shall be included in all
#copies or substantial portions of the Software.
#THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
#IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
#FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
#AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
#LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
#OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
#SOFTWARE.
##################################]
# helpers
def pair(t):
return t if isinstance(t, tuple) else (t, t)
def posemb_sincos_2d(h, w, dim, temperature: int = 10000, dtype = torch.float32):
y, x = torch.meshgrid(torch.arange(h), torch.arange(w), indexing="ij")
assert (dim % 4) == 0, "feature dimension must be multiple of 4 for sincos emb"
omega = torch.arange(dim // 4) / (dim // 4 - 1)
omega = 1.0 / (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)
# 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):
### Here is my MOE modification
def __init__(self, dim, depth, heads, dim_head, mlp_dim, num_experts):
super().__init__()
self.norm = nn.LayerNorm(dim)
self.layers = nn.ModuleList([])
for _ in range(depth):
if _ % 2 == 0: #make every other FNN an expert layer.
self.layers.append(nn.ModuleList([
Attention(dim, heads = heads, dim_head = dim_head),
FeedForward(dim, mlp_dim)
]))
else:
self.layers.append(nn.ModuleList([
Attention(dim, heads = heads, dim_head = dim_head),
SparseMoEBlock(
MoE(dim = dim,
num_experts = num_experts,
gating_top_n = 2,
threshold_train = 0.2,
threshold_eval = 0.2,
capacity_factor_train = 1.25,
capacity_factor_eval = 2.,
balance_loss_coef = 1e-2,
router_z_loss_coef = 1e-3,
),
add_ff_before = True,
add_ff_after = True
)
]))
def forward(self, x):
for attne, ff in self.layers:
x = attne(x) + x
try:
x = ff(x) + x
except:
x = ff(x)[0]+x # I won't bother returning aux_loss... probably a bad idea
return self.norm(x)
class SimpleViTMIX(nn.Module):
def __init__(self, *, image_size, patch_size, num_classes, dim, depth, heads, mlp_dim, channels = 3, dim_head = 64, num_experts = 12):
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.'
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.LayerNorm(patch_dim),
nn.Linear(patch_dim, dim),
nn.LayerNorm(dim),
)
self.pos_embedding = posemb_sincos_2d(
h = image_height // patch_height,
w = image_width // patch_width,
dim = dim,
)
self.transformer = Transformer(dim, depth, heads, dim_head, mlp_dim, num_experts)
self.pool = "mean"
self.to_latent = nn.Identity()
self.linear_head = nn.Linear(dim, num_classes)
def forward(self, img):
device = img.device
x = self.to_patch_embedding(img)
x += self.pos_embedding.to(device, dtype=x.dtype)
x = self.transformer(x)
x = x.mean(dim = 1)
x = self.to_latent(x)
return self.linear_head(x)
class ViTMixModel(PreTrainedModel):
config_class = ViTMixConfig
def __init__(self, config):
super().__init__(config)
self.model = SimpleViTMIX(
image_size = config.image_size,
patch_size = config.patch_size,
num_classes = config.num_classes,
dim = config.dim,
depth = config.depth,
heads = config.heads,
mlp_dim = config.mlp_dim,
num_experts = config.num_experts
)
def forward(self,tensor, labels = None):
logits = self.model(tensor)
if labels is not None:
loss = torch.nn.cross_entropy(logits, labels)
return {"loss": loss, "logits": logits}
return {"logits": logits} |