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
|
import os
import numpy as np
import PIL
import torch
from PIL import Image
from torch.utils.data import Dataset, DataLoader
from torchvision import transforms
import random
import tqdm
from modules import devices, shared
import re
from ldm.modules.distributions.distributions import DiagonalGaussianDistribution
re_numbers_at_start = re.compile(r"^[-\d]+\s*")
class DatasetEntry:
def __init__(self, filename=None, filename_text=None, latent_dist=None, latent_sample=None, cond=None, cond_text=None, pixel_values=None):
self.filename = filename
self.filename_text = filename_text
self.latent_dist = latent_dist
self.latent_sample = latent_sample
self.cond = cond
self.cond_text = cond_text
self.pixel_values = pixel_values
class PersonalizedBase(Dataset):
def __init__(self, data_root, width, height, repeats, flip_p=0.5, placeholder_token="*", model=None, cond_model=None, device=None, template_file=None, include_cond=False, batch_size=1, gradient_step=1, shuffle_tags=False, tag_drop_out=0, latent_sampling_method='once'):
re_word = re.compile(shared.opts.dataset_filename_word_regex) if len(shared.opts.dataset_filename_word_regex) > 0 else None
self.placeholder_token = placeholder_token
self.width = width
self.height = height
self.flip = transforms.RandomHorizontalFlip(p=flip_p)
self.dataset = []
with open(template_file, "r") as file:
lines = [x.strip() for x in file.readlines()]
self.lines = lines
assert data_root, 'dataset directory not specified'
assert os.path.isdir(data_root), "Dataset directory doesn't exist"
assert os.listdir(data_root), "Dataset directory is empty"
self.image_paths = [os.path.join(data_root, file_path) for file_path in os.listdir(data_root)]
self.shuffle_tags = shuffle_tags
self.tag_drop_out = tag_drop_out
print("Preparing dataset...")
for path in tqdm.tqdm(self.image_paths):
if shared.state.interrupted:
raise Exception("inturrupted")
try:
image = Image.open(path).convert('RGB').resize((self.width, self.height), PIL.Image.BICUBIC)
except Exception:
continue
text_filename = os.path.splitext(path)[0] + ".txt"
filename = os.path.basename(path)
if os.path.exists(text_filename):
with open(text_filename, "r", encoding="utf8") as file:
filename_text = file.read()
else:
filename_text = os.path.splitext(filename)[0]
filename_text = re.sub(re_numbers_at_start, '', filename_text)
if re_word:
tokens = re_word.findall(filename_text)
filename_text = (shared.opts.dataset_filename_join_string or "").join(tokens)
npimage = np.array(image).astype(np.uint8)
npimage = (npimage / 127.5 - 1.0).astype(np.float32)
torchdata = torch.from_numpy(npimage).permute(2, 0, 1).to(device=device, dtype=torch.float32)
latent_sample = None
with torch.autocast("cuda"):
latent_dist = model.encode_first_stage(torchdata.unsqueeze(dim=0))
if latent_sampling_method == "once" or (latent_sampling_method == "deterministic" and not isinstance(latent_dist, DiagonalGaussianDistribution)):
latent_sample = model.get_first_stage_encoding(latent_dist).squeeze().to(devices.cpu)
latent_sampling_method = "once"
entry = DatasetEntry(filename=path, filename_text=filename_text, latent_sample=latent_sample)
elif latent_sampling_method == "deterministic":
# Works only for DiagonalGaussianDistribution
latent_dist.std = 0
latent_sample = model.get_first_stage_encoding(latent_dist).squeeze().to(devices.cpu)
entry = DatasetEntry(filename=path, filename_text=filename_text, latent_sample=latent_sample)
elif latent_sampling_method == "random":
entry = DatasetEntry(filename=path, filename_text=filename_text, latent_dist=latent_dist)
if not (self.tag_drop_out != 0 or self.shuffle_tags):
entry.cond_text = self.create_text(filename_text)
if include_cond and not (self.tag_drop_out != 0 or self.shuffle_tags):
with torch.autocast("cuda"):
entry.cond = cond_model([entry.cond_text]).to(devices.cpu).squeeze(0)
# elif not include_cond:
# _, _, _, _, hijack_fixes, token_count = cond_model.process_text([entry.cond_text])
# max_n = token_count // 75
# index_list = [ [] for _ in range(max_n + 1) ]
# for n, (z, _) in hijack_fixes[0]:
# index_list[n].append(z)
# with torch.autocast("cuda"):
# entry.cond = cond_model([entry.cond_text]).to(devices.cpu).squeeze(0)
# entry.emb_index = index_list
self.dataset.append(entry)
del torchdata
del latent_dist
del latent_sample
self.length = len(self.dataset)
assert self.length > 0, "No images have been found in the dataset."
self.batch_size = min(batch_size, self.length)
self.gradient_step = min(gradient_step, self.length // self.batch_size)
self.latent_sampling_method = latent_sampling_method
def create_text(self, filename_text):
text = random.choice(self.lines)
text = text.replace("[name]", self.placeholder_token)
tags = filename_text.split(',')
if self.tag_drop_out != 0:
tags = [t for t in tags if random.random() > self.tag_drop_out]
if self.shuffle_tags:
random.shuffle(tags)
text = text.replace("[filewords]", ','.join(tags))
return text
def __len__(self):
return self.length
def __getitem__(self, i):
entry = self.dataset[i]
if self.tag_drop_out != 0 or self.shuffle_tags:
entry.cond_text = self.create_text(entry.filename_text)
if self.latent_sampling_method == "random":
entry.latent_sample = shared.sd_model.get_first_stage_encoding(entry.latent_dist)
return entry
class PersonalizedDataLoader(DataLoader):
def __init__(self, *args, **kwargs):
super(PersonalizedDataLoader, self).__init__(shuffle=True, drop_last=True, *args, **kwargs)
self.collate_fn = collate_wrapper
class BatchLoader:
def __init__(self, data):
self.cond_text = [entry.cond_text for entry in data]
self.cond = [entry.cond for entry in data]
self.latent_sample = torch.stack([entry.latent_sample for entry in data]).squeeze(1)
def pin_memory(self):
self.latent_sample = self.latent_sample.pin_memory()
return self
def collate_wrapper(batch):
return BatchLoader(batch)
|
