File size: 10,875 Bytes
5e373a9 |
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 |
# This code is built from the Stable Diffusion repository: https://github.com/CompVis/stable-diffusion, and
# Paint-by-Example repo https://github.com/Fantasy-Studio/Paint-by-Example
# Copyright (c) 2022 Robin Rombach and Patrick Esser and contributors.
# CreativeML Open RAIL-M
#
# ==========================================================================================
#
# Adobe’s modifications are Copyright 2024 Adobe Research. All rights reserved.
# Adobe’s modifications are licensed under the Adobe Research License. To view a copy of the license, visit
# LICENSE.md.
#
# ==========================================================================================
import torch
import torch.nn as nn
from functools import partial
import clip
from einops import rearrange, repeat
from transformers import CLIPTokenizer, CLIPTextModel,CLIPVisionModel,CLIPModel
import kornia
from ldm.modules.x_transformer import Encoder, TransformerWrapper # TODO: can we directly rely on lucidrains code and simply add this as a reuirement? --> test
from .xf import LayerNorm, Transformer
import math
class AbstractEncoder(nn.Module):
def __init__(self):
super().__init__()
def encode(self, *args, **kwargs):
raise NotImplementedError
class ClassEmbedder(nn.Module):
def __init__(self, embed_dim, n_classes=1000, key='class'):
super().__init__()
self.key = key
self.embedding = nn.Embedding(n_classes, embed_dim)
def forward(self, batch, key=None):
if key is None:
key = self.key
# this is for use in crossattn
c = batch[key][:, None]
c = self.embedding(c)
return c
class TransformerEmbedder(AbstractEncoder):
"""Some transformer encoder layers"""
def __init__(self, n_embed, n_layer, vocab_size, max_seq_len=77, device="cuda"):
super().__init__()
self.device = device
self.transformer = TransformerWrapper(num_tokens=vocab_size, max_seq_len=max_seq_len,
attn_layers=Encoder(dim=n_embed, depth=n_layer))
def forward(self, tokens):
tokens = tokens.to(self.device) # meh
z = self.transformer(tokens, return_embeddings=True)
return z
def encode(self, x):
return self(x)
class BERTTokenizer(AbstractEncoder):
""" Uses a pretrained BERT tokenizer by huggingface. Vocab size: 30522 (?)"""
def __init__(self, device="cuda", vq_interface=True, max_length=77):
super().__init__()
from transformers import BertTokenizerFast # TODO: add to reuquirements
self.tokenizer = BertTokenizerFast.from_pretrained("bert-base-uncased")
self.device = device
self.vq_interface = vq_interface
self.max_length = max_length
def forward(self, text):
batch_encoding = self.tokenizer(text, truncation=True, max_length=self.max_length, return_length=True,
return_overflowing_tokens=False, padding="max_length", return_tensors="pt")
tokens = batch_encoding["input_ids"].to(self.device)
return tokens
@torch.no_grad()
def encode(self, text):
tokens = self(text)
if not self.vq_interface:
return tokens
return None, None, [None, None, tokens]
def decode(self, text):
return text
class BERTEmbedder(AbstractEncoder):
"""Uses the BERT tokenizr model and add some transformer encoder layers"""
def __init__(self, n_embed, n_layer, vocab_size=30522, max_seq_len=77,
device="cuda",use_tokenizer=True, embedding_dropout=0.0):
super().__init__()
self.use_tknz_fn = use_tokenizer
if self.use_tknz_fn:
self.tknz_fn = BERTTokenizer(vq_interface=False, max_length=max_seq_len)
self.device = device
self.transformer = TransformerWrapper(num_tokens=vocab_size, max_seq_len=max_seq_len,
attn_layers=Encoder(dim=n_embed, depth=n_layer),
emb_dropout=embedding_dropout)
def forward(self, text):
if self.use_tknz_fn:
tokens = self.tknz_fn(text)#.to(self.device)
else:
tokens = text
z = self.transformer(tokens, return_embeddings=True)
return z
def encode(self, text):
# output of length 77
return self(text)
class FrozenCLIPEmbedder(AbstractEncoder):
"""Uses the CLIP transformer encoder for text (from Hugging Face)"""
def __init__(self, version="openai/clip-vit-large-patch14", device="cuda", max_length=77):
super().__init__()
self.tokenizer = CLIPTokenizer.from_pretrained(version)
self.transformer = CLIPTextModel.from_pretrained(version)
self.device = device
self.max_length = max_length
self.freeze()
def freeze(self):
self.transformer = self.transformer.eval()
for param in self.parameters():
param.requires_grad = False
def forward(self, text):
batch_encoding = self.tokenizer(text, truncation=True, max_length=self.max_length, return_length=True,
return_overflowing_tokens=False, padding="max_length", return_tensors="pt")
tokens = batch_encoding["input_ids"].to(self.device)
outputs = self.transformer(input_ids=tokens)
z = outputs.last_hidden_state
return z
def encode(self, text):
return self(text)
class SpatialRescaler(nn.Module):
def __init__(self,
n_stages=1,
method='bilinear',
multiplier=0.5,
in_channels=3,
out_channels=None,
bias=False):
super().__init__()
self.n_stages = n_stages
assert self.n_stages >= 0
assert method in ['nearest','linear','bilinear','trilinear','bicubic','area']
self.multiplier = multiplier
self.interpolator = partial(torch.nn.functional.interpolate, mode=method)
self.remap_output = out_channels is not None
if self.remap_output:
print(f'Spatial Rescaler mapping from {in_channels} to {out_channels} channels after resizing.')
self.channel_mapper = nn.Conv2d(in_channels,out_channels,1,bias=bias)
def forward(self,x):
for stage in range(self.n_stages):
x = self.interpolator(x, scale_factor=self.multiplier)
if self.remap_output:
x = self.channel_mapper(x)
return x
def encode(self, x):
return self(x)
class FrozenCLIPTextEmbedder(nn.Module):
"""
Uses the CLIP transformer encoder for text.
"""
def __init__(self, version='ViT-L/14', device="cuda", max_length=77, n_repeat=1, normalize=True):
super().__init__()
self.model, _ = clip.load(version, jit=False, device="cpu")
self.device = device
self.max_length = max_length
self.n_repeat = n_repeat
self.normalize = normalize
def freeze(self):
self.model = self.model.eval()
for param in self.parameters():
param.requires_grad = False
def forward(self, text):
tokens = clip.tokenize(text).to(self.device)
z = self.model.encode_text(tokens)
if self.normalize:
z = z / torch.linalg.norm(z, dim=1, keepdim=True)
return z
def encode(self, text):
z = self(text)
if z.ndim==2:
z = z[:, None, :]
z = repeat(z, 'b 1 d -> b k d', k=self.n_repeat)
return z
class FrozenCLIPImageEmbedder(AbstractEncoder):
"""Uses the CLIP transformer encoder for text (from Hugging Face)"""
def __init__(self, version="openai/clip-vit-large-patch14"):
super().__init__()
self.transformer = CLIPVisionModel.from_pretrained(version)
self.final_ln = LayerNorm(1024)
self.mapper = Transformer(
1,
1024,
5,
1,
)
self.freeze()
def freeze(self):
self.transformer = self.transformer.eval()
for param in self.parameters():
param.requires_grad = False
for param in self.mapper.parameters():
param.requires_grad = True
for param in self.final_ln.parameters():
param.requires_grad = True
def forward(self, image):
outputs = self.transformer(pixel_values=image)
z = outputs.pooler_output
z = z.unsqueeze(1)
z = self.mapper(z)
z = self.final_ln(z)
return z
def encode(self, image):
return self(image)
class DINOEmbedder(AbstractEncoder):
"""Uses the CLIP transformer encoder for text (from Hugging Face)"""
def __init__(self, dino_version): # small, large, huge, gigantic
super().__init__()
assert dino_version in ['small', 'big', 'large', 'huge']
letter_map = {
'small': 's',
'big': 'b',
'large': 'l',
'huge': 'g'
}
self.final_ln = LayerNorm(32) # unused -- remove later
self.mapper = LayerNorm(32) # unused -- remove later
# embedding_sizes = {
# 'small': 384,
# 'big': 768,
# 'large': 1024,
# 'huge': 1536
# }
# embedding_size = embedding_sizes[dino_version]
letter = letter_map[dino_version]
# self.transformer = CLIPVisionModel.from_pretrained(version)
self.dino_model = torch.hub.load('facebookresearch/dinov2', f'dinov2_vit{letter}14_reg').cuda()
self.freeze()
def freeze(self):
for param in self.parameters():
param.requires_grad = False
def forward(self, image):
with torch.no_grad():
outputs = self.dino_model.forward_features(image)
patch_tokens = outputs['x_norm_patchtokens']
global_token = outputs['x_norm_clstoken'].unsqueeze(1)
features = torch.concat([patch_tokens, global_token], dim=1)
return torch.zeros_like(features)
def encode(self, image):
return self(image)
class FixedVector(AbstractEncoder):
"""Uses the CLIP transformer encoder for text (from Hugging Face)"""
def __init__(self): # small, large, huge, gigantic
super().__init__()
self.final_ln = LayerNorm(32)
self.mapper = LayerNorm(32)
self.fixed_vector = nn.Parameter(torch.randn((1,1,768)), requires_grad=True).cuda()
def forward(self, image):
return self.fixed_vector.repeat(image.shape[0],1,1).to(image.device) * 0.0
def encode(self, image):
return self(image)
if __name__ == "__main__":
from ldm.util import count_params
model = FrozenCLIPEmbedder()
count_params(model, verbose=True) |