File size: 20,908 Bytes
c8c00f0 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 | # Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
# This source code is licensed under the license found in the
# LICENSE file in the root directory of this source tree.
# --------------------------------------------------------
# References:
# GLIDE: https://github.com/openai/glide-text2im
# MAE: https://github.com/facebookresearch/mae/blob/main/models_mae.py
# --------------------------------------------------------
import torch
import torch.nn as nn
import numpy as np
import math
from timm.models.vision_transformer import PatchEmbed, Attention, Mlp
from torch import einsum
from einops import rearrange, repeat
from autoencoder import *
def modulate(x, shift, scale):
return x * (1 + scale.unsqueeze(1)) + shift.unsqueeze(1)
#################################################################################
# Embedding Layers for Timesteps and Class Labels #
#################################################################################
class TimestepEmbedder(nn.Module):
"""
Embeds scalar timesteps into vector representations.
"""
def __init__(self, hidden_size, frequency_embedding_size=256):
super().__init__()
self.mlp = nn.Sequential(
nn.Linear(frequency_embedding_size, hidden_size, bias=True),
nn.SiLU(),
nn.Linear(hidden_size, hidden_size, bias=True),
)
self.frequency_embedding_size = frequency_embedding_size
@staticmethod
def timestep_embedding(t, dim, max_period=10000):
"""
Create sinusoidal timestep embeddings.
:param t: a 1-D Tensor of N indices, one per batch element.
These may be fractional.
:param dim: the dimension of the output.
:param max_period: controls the minimum frequency of the embeddings.
:return: an (N, D) Tensor of positional embeddings.
"""
# https://github.com/openai/glide-text2im/blob/main/glide_text2im/nn.py
half = dim // 2
freqs = torch.exp(
-math.log(max_period) * torch.arange(start=0, end=half, dtype=torch.float32) / half
).to(device=t.device)
args = t[:, None].float() * freqs[None]
embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
if dim % 2:
embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
return embedding
def forward(self, t):
t_freq = self.timestep_embedding(t, self.frequency_embedding_size)
t_emb = self.mlp(t_freq)
return t_emb
#################################################################################
# Core DiT Model #
#################################################################################
class CrossAttention(nn.Module):
def __init__(self, query_dim, heads=8, dropout=0.):
super().__init__()
dim_head = query_dim / heads
self.scale = dim_head ** -0.5
self.heads = heads
self.to_q = nn.Linear(query_dim, query_dim, bias=True)
self.to_k = nn.Linear(query_dim, query_dim, bias=True)
self.to_v = nn.Linear(query_dim, query_dim, bias=True)
def forward(self, x, context=None):
h = self.heads
q = self.to_q(x)
k = self.to_k(context).unsqueeze(1)
v = self.to_v(context).unsqueeze(1)
q, k, v = map(lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v))
sim = einsum('b i d, b j d -> b i j', q, k) * self.scale
# attention, what we cannot get enough of
attn = sim.softmax(dim=-2)
out = einsum('b i j, b j d -> b i d', attn, v)
out = rearrange(out, '(b h) n d -> b n (h d)', h=h)
attn_out = rearrange(attn, '(b h) n d -> b n (h d)', h=h)
return out, attn_out
class Cross_Norm(nn.Module):
def __init__(self, hidden_size, num_heads):
super().__init__()
self.norm = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
self.cross_attention = CrossAttention(hidden_size, heads=num_heads)
def forward(self, x, c):
x = self.norm(x)
x = self.cross_attention(x, c)
return x
class DiTBlock(nn.Module):
"""
A DiT block with adaptive layer norm zero (adaLN-Zero) conditioning.
"""
def __init__(self, hidden_size, num_heads, mlp_ratio=4.0, **block_kwargs):
super().__init__()
self.norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
self.attn = Attention(hidden_size, num_heads=num_heads, qkv_bias=True, **block_kwargs)
self.norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
mlp_hidden_dim = int(hidden_size * mlp_ratio)
approx_gelu = lambda: nn.GELU()
self.mlp = Mlp(in_features=hidden_size, hidden_features=mlp_hidden_dim, act_layer=approx_gelu, drop=0)
self.adaLN_modulation = nn.Sequential(
nn.SiLU(),
nn.Linear(hidden_size, 6 * hidden_size, bias=True)
)
def forward(self, x, c):
shift_msa, scale_msa, gate_msa, shift_mlp, scale_mlp, gate_mlp = self.adaLN_modulation(c).chunk(6, dim=1)
x = x + gate_msa.unsqueeze(1) * self.attn(modulate(self.norm1(x), shift_msa, scale_msa))
x = x + gate_mlp.unsqueeze(1) * self.mlp(modulate(self.norm2(x), shift_mlp, scale_mlp))
return x
class FinalLayer(nn.Module):
"""
The final layer of DiT.
"""
def __init__(self, hidden_size, patch_size, out_channels):
super().__init__()
self.norm_final = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
self.linear = nn.Linear(hidden_size, patch_size * patch_size * out_channels, bias=True)
self.adaLN_modulation = nn.Sequential(
nn.SiLU(),
nn.Linear(hidden_size, 2 * hidden_size, bias=True)
)
def forward(self, x, c):
shift, scale = self.adaLN_modulation(c).chunk(2, dim=1)
x = modulate(self.norm_final(x), shift, scale)
x = self.linear(x)
return x
class temp_Adaptive_Mask(nn.Module):
def __init__(self, hidden_size, patch_size, out_channels):
super().__init__()
self.out_channels = out_channels
self.patch_size = patch_size
self.norm = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
self.linear = nn.Linear(hidden_size, patch_size * patch_size * out_channels, bias=True)
self.mlp = nn.Sequential(
nn.SiLU(),
nn.Linear(hidden_size, 2 * hidden_size, bias=True),
nn.Linear(2 * hidden_size, hidden_size, bias=True)
)
def forward(self, x):
x = self.norm(x)
x = self.mlp(x)
x = self.linear(x)
h = w = int(x.shape[1] ** 0.5)
assert h * w == x.shape[1]
c = self.out_channels
p = self.patch_size
x = x.reshape(shape=(x.shape[0], h, w, p, p, c))
x = torch.einsum('nhwpqc->nchpwq', x)
x = x.reshape(shape=(x.shape[0], c, h * p, h * p))
return x
class DiT(nn.Module):
"""
Diffusion model with a Transformer backbone.
"""
def __init__(
self,
input_size=32,
patch_size=2,
in_channels=4,
hidden_size=1152,
depth=28,
num_heads=16,
mlp_ratio=4.0,
class_dropout_prob=0.1,
num_classes=1000,
learn_sigma=True,
):
super().__init__()
self.learn_sigma = learn_sigma
self.in_channels = in_channels
self.out_channels = in_channels * 2 if learn_sigma else in_channels
self.patch_size = patch_size
self.num_heads = num_heads
self.x_embedder = PatchEmbed(input_size, patch_size, in_channels, hidden_size, bias=True)
self.t_embedder = TimestepEmbedder(hidden_size)
self.y_embedders = nn.Linear(1024, 1152)
num_patches = self.x_embedder.num_patches
self.pos_embed = nn.Parameter(torch.zeros(1, num_patches, hidden_size), requires_grad=False)
self.blocks = nn.ModuleList([
DiTBlock(hidden_size, num_heads, mlp_ratio=mlp_ratio) for _ in range(depth)
])
self.cross_defect = nn.ModuleList([
Cross_Norm(hidden_size, num_heads) for _ in range(10)
])
self.adapt_mask = temp_Adaptive_Mask(num_heads*10, patch_size, in_channels)
self.final_layer = FinalLayer(hidden_size, patch_size, self.out_channels)
self.initialize_weights()
def initialize_weights(self):
# Initialize transformer layers:
def _basic_init(module):
if isinstance(module, nn.Linear):
torch.nn.init.xavier_uniform_(module.weight)
if module.bias is not None:
nn.init.constant_(module.bias, 0)
self.apply(_basic_init)
# Initialize (and freeze) pos_embed by sin-cos embedding:
pos_embed = get_2d_sincos_pos_embed(self.pos_embed.shape[-1], int(self.x_embedder.num_patches ** 0.5))
self.pos_embed.data.copy_(torch.from_numpy(pos_embed).float().unsqueeze(0))
# Initialize patch_embed like nn.Linear (instead of nn.Conv2d):
w = self.x_embedder.proj.weight.data
nn.init.xavier_uniform_(w.view([w.shape[0], -1]))
nn.init.constant_(self.x_embedder.proj.bias, 0)
# Initialize timestep embedding MLP:
nn.init.normal_(self.t_embedder.mlp[0].weight, std=0.02)
nn.init.normal_(self.t_embedder.mlp[2].weight, std=0.02)
# Zero-out adaLN modulation layers in DiT blocks:
for block in self.blocks:
nn.init.constant_(block.adaLN_modulation[-1].weight, 0)
nn.init.constant_(block.adaLN_modulation[-1].bias, 0)
# Zero-out output layers:
nn.init.constant_(self.final_layer.adaLN_modulation[-1].weight, 0)
nn.init.constant_(self.final_layer.adaLN_modulation[-1].bias, 0)
nn.init.constant_(self.final_layer.linear.weight, 0)
nn.init.constant_(self.final_layer.linear.bias, 0)
def unpatchify(self, x):
"""
x: (N, T, patch_size**2 * C)
imgs: (N, H, W, C)
"""
c = self.out_channels
p = self.x_embedder.patch_size[0]
h = w = int(x.shape[1] ** 0.5)
assert h * w == x.shape[1]
x = x.reshape(shape=(x.shape[0], h, w, p, p, c))
x = torch.einsum('nhwpqc->nchpwq', x)
imgs = x.reshape(shape=(x.shape[0], c, h * p, h * p))
return imgs
def forward(self, x, t, y):
"""
Forward pass of DiT.
x: (N, C, H, W) tensor of spatial inputs (images or latent representations of images)
t: (N,) tensor of diffusion timesteps
y: (N,) tensor of class labels
"""
x = self.x_embedder(x) + self.pos_embed # (N, T, D), where T = H * W / patch_size ** 2
t = self.t_embedder(t) # (N, D)
y_defect = self.y_embedders(y[0])
y_class = self.y_embedders(y[1]) # (N, D)
y_all = self.y_embedders(y[2])
att_map = []
loss_att = 0
for i in range(28):
block = self.blocks[i]
if i < 10:
c = t + y_class
x = block(x, c) # (N, T, D)
elif i < 20:
cross = self.cross_defect[i - 10]
c = t + y_defect
shift_msa, scale_msa, gate_msa, shift_mlp, scale_mlp, gate_mlp = block.adaLN_modulation(c).chunk(6,
dim=1)
x = x + gate_msa.unsqueeze(1) * block.attn(modulate(block.norm1(x), shift_msa, scale_msa))
cross_att, att_weight = cross(x, c)
loss_att += att_weight
att_map.append(att_weight)
x = x + cross_att
x = x + gate_mlp.unsqueeze(1) * block.mlp(modulate(block.norm2(x), shift_mlp, scale_mlp))
elif i < 28:
c = t + y_all
x = block(x, c)
x = self.final_layer(x, c) # (N, T, patch_size ** 2 * out_channels)
x = self.unpatchify(x) # (N, out_channels, H, W)
att_map = torch.cat(att_map, dim=-1)
att_mask = self.adapt_mask(att_map)
return x, att_mask, loss_att.resize(x.shape[0], x.shape[2]//2, x.shape[3]//2, 16).mean(dim=-1)
def forward_free_2(self, x, t, y, mask_temp=None):
"""
Forward pass of DiT.
x: (N, C, H, W) tensor of spatial inputs (images or latent representations of images)
t: (N,) tensor of diffusion timesteps
y: (N,) tensor of class labels
"""
x = self.x_embedder(x) + self.pos_embed # (N, T, D), where T = H * W / patch_size ** 2
t = self.t_embedder(t) # (N, D)
y_defect = self.y_embedders(torch.cat([y[0][0], y[1][0]], dim=0))
y_class = self.y_embedders(torch.cat([y[0][1], y[1][1]], dim=0)) # (N, D)
y_all = self.y_embedders(torch.cat([y[0][2], y[1][2]], dim=0))
att_map = []
loss_att = 0
for i in range(28):
block = self.blocks[i]
if i < 10:
c = t + y_class
x = block(x, c) # (N, T, D)
elif i < 20:
cross = self.cross_defect[i - 10]
c = t + y_defect
shift_msa, scale_msa, gate_msa, shift_mlp, scale_mlp, gate_mlp = block.adaLN_modulation(c).chunk(6,
dim=1)
x = x + gate_msa.unsqueeze(1) * block.attn(modulate(block.norm1(x), shift_msa, scale_msa))
cross_att, att_weight = cross(x, c)
att_map.append(att_weight)
loss_att+=att_weight
x = x + cross_att
x = x + gate_mlp.unsqueeze(1) * block.mlp(modulate(block.norm2(x), shift_mlp, scale_mlp))
elif i < 28:
c = t + y_all
x = block(x, c)
x = self.final_layer(x, c) # (N, T, patch_size ** 2 * out_channels)
x = self.unpatchify(x) # (N, out_channels, H, W)
att_map = torch.cat(att_map, dim=-1)
att_mask = self.adapt_mask(att_map)
return x, att_mask, loss_att
def forward_with_cfg_2(self, x, t, y, cfg_scale):
"""
Forward pass of DiT, but also batches the unconditional forward pass for classifier-free guidance.
"""
# https://github.com/openai/glide-text2im/blob/main/notebooks/text2im.ipynb
half = x[: len(x) // 2]
combined = torch.cat([half, half], dim=0)
model_out, mask, _ = self.forward_free_2(combined, t, y)
# For exact reproducibility reasons, we apply classifier-free guidance on only
# three channels by default. The standard approach to cfg applies it to all channels.
# This can be done by uncommenting the following line and commenting-out the line following that.
# eps, rest = model_out[:, :self.in_channels], model_out[:, self.in_channels:]
eps, rest = model_out[:, :3], model_out[:, 3:]
cond_eps, uncond_eps = torch.split(eps, len(eps) // 2, dim=0)
half_eps = uncond_eps + cfg_scale * (cond_eps - uncond_eps)
eps = torch.cat([half_eps, half_eps], dim=0)
return torch.cat([eps, rest], dim=1), mask, _
def forward_free_3(self, x, t, y, mask_temp=None):
"""
Forward pass of DiT.
x: (N, C, H, W) tensor of spatial inputs (images or latent representations of images)
t: (N,) tensor of diffusion timesteps
y: (N,) tensor of class labels
"""
x = self.x_embedder(x) + self.pos_embed # (N, T, D), where T = H * W / patch_size ** 2
t = self.t_embedder(t) # (N, D)
y_defect = self.y_embedders(torch.cat([y[0][0], y[1][0], y[2][0]], dim=0))
y_class = self.y_embedders(torch.cat([y[0][1], y[1][1], y[2][1]], dim=0)) # (N, D)
y_all = self.y_embedders(torch.cat([y[0][2], y[1][2], y[2][2]], dim=0))
att_map = []
att_loss = 0
for i in range(28):
block = self.blocks[i]
if i < 10:
c = t + y_class
x = block(x, c) # (N, T, D)
elif i < 20:
cross = self.cross_defect[i - 10]
c = t + y_defect
shift_msa, scale_msa, gate_msa, shift_mlp, scale_mlp, gate_mlp = block.adaLN_modulation(c).chunk(6,
dim=1)
x = x + gate_msa.unsqueeze(1) * block.attn(modulate(block.norm1(x), shift_msa, scale_msa))
cross_att, att_weight = cross(x, c)
att_map.append(att_weight)
att_loss += att_weight
x = x + cross_att
x = x + gate_mlp.unsqueeze(1) * block.mlp(modulate(block.norm2(x), shift_mlp, scale_mlp))
elif i < 28:
c = t + y_all
x = block(x, c)
x = self.final_layer(x, c) # (N, T, patch_size ** 2 * out_channels)
x = self.unpatchify(x) # (N, out_channels, H, W)
att_map = torch.cat(att_map, dim=-1)
att_mask = self.adapt_mask(att_map)
return x, att_mask, att_loss
def forward_with_cfg_3(self, x, t, y, cfg_scale):
"""
Forward pass of DiT, but also batches the unconditional forward pass for classifier-free guidance.
"""
# https://github.com/openai/glide-text2im/blob/main/notebooks/text2im.ipynb
half = x[: len(x) // 3]
combined = torch.cat([half, half, half], dim=0)
model_out, mask, _ = self.forward_free_3(combined, t, y)
# For exact reproducibility reasons, we apply classifier-free guidance on only
# three channels by default. The standard approach to cfg applies it to all channels.
# This can be done by uncommenting the following line and commenting-out the line following that.
# eps, rest = model_out[:, :self.in_channels], model_out[:, self.in_channels:]
eps, rest = model_out[:, :3], model_out[:, 3:]
cond_eps, uncond_eps_defect, uncond_eps = torch.split(eps, len(eps) // 3, dim=0)
half_eps = uncond_eps + cfg_scale * (cond_eps - uncond_eps_defect) + cfg_scale * (uncond_eps_defect - uncond_eps)
eps = torch.cat([half_eps, half_eps, half_eps], dim=0)
return torch.cat([eps, rest], dim=1), mask, _
#################################################################################
# Sine/Cosine Positional Embedding Functions #
#################################################################################
# https://github.com/facebookresearch/mae/blob/main/util/pos_embed.py
def get_2d_sincos_pos_embed(embed_dim, grid_size, cls_token=False, extra_tokens=0):
"""
grid_size: int of the grid height and width
return:
pos_embed: [grid_size*grid_size, embed_dim] or [1+grid_size*grid_size, embed_dim] (w/ or w/o cls_token)
"""
grid_h = np.arange(grid_size, dtype=np.float32)
grid_w = np.arange(grid_size, dtype=np.float32)
grid = np.meshgrid(grid_w, grid_h) # here w goes first
grid = np.stack(grid, axis=0)
grid = grid.reshape([2, 1, grid_size, grid_size])
pos_embed = get_2d_sincos_pos_embed_from_grid(embed_dim, grid)
if cls_token and extra_tokens > 0:
pos_embed = np.concatenate([np.zeros([extra_tokens, embed_dim]), pos_embed], axis=0)
return pos_embed
def get_2d_sincos_pos_embed_from_grid(embed_dim, grid):
assert embed_dim % 2 == 0
# use half of dimensions to encode grid_h
emb_h = get_1d_sincos_pos_embed_from_grid(embed_dim // 2, grid[0]) # (H*W, D/2)
emb_w = get_1d_sincos_pos_embed_from_grid(embed_dim // 2, grid[1]) # (H*W, D/2)
emb = np.concatenate([emb_h, emb_w], axis=1) # (H*W, D)
return emb
def get_1d_sincos_pos_embed_from_grid(embed_dim, pos):
"""
embed_dim: output dimension for each position
pos: a list of positions to be encoded: size (M,)
out: (M, D)
"""
assert embed_dim % 2 == 0
omega = np.arange(embed_dim // 2, dtype=np.float64)
omega /= embed_dim / 2.
omega = 1. / 10000**omega # (D/2,)
pos = pos.reshape(-1) # (M,)
out = np.einsum('m,d->md', pos, omega) # (M, D/2), outer product
emb_sin = np.sin(out) # (M, D/2)
emb_cos = np.cos(out) # (M, D/2)
emb = np.concatenate([emb_sin, emb_cos], axis=1) # (M, D)
return emb
|