-
Notifications
You must be signed in to change notification settings - Fork 0
/
data_utils.py
330 lines (272 loc) · 11 KB
/
data_utils.py
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
from __future__ import annotations
import os
from copy import copy
import cv2
from datasets import load_dataset
from torch.utils import data
from torchvision.transforms import (CenterCrop,
Compose,
Normalize,
RandomHorizontalFlip,
RandomResizedCrop,
Resize,
ToTensor)
skip_exec = True
def prepare_datasets(model_name: str, task_name: str, data_name: str, tokenizer, cache_dir: str, eval_key: str = 'val'):
if task_name == 'glue':
return prepare_datasets_glue(model_name, data_name, tokenizer, cache_dir, eval_key)
elif task_name == 'img_class':
if 'cifar' in data_name:
return prepare_datasets_cifar(model_name, data_name, tokenizer, cache_dir, eval_key)
elif data_name == 'tinyimagenet':
return prepare_datasets_tinyimagenet(model_name, data_name, tokenizer, cache_dir, eval_key)
else:
raise NotImplementedError
elif task_name == 'img_seg':
if 'cityscapes' in data_name:
return prepare_datasets_cityscapes(model_name, data_name, tokenizer, cache_dir, eval_key)
elif 'kitti' in data_name:
return prepare_datasets_kitti(model_name, data_name, tokenizer, cache_dir, eval_key)
else:
raise NotImplementedError
def prepare_datasets_glue(model_name: str, data_name: str, tokenizer, cache_dir: str, eval_key: str = 'val'):
task_to_keys = {
'cola': ('sentence', None),
'mnli': ('premise', 'hypothesis'),
'mrpc': ('sentence1', 'sentence2'),
'qnli': ('question', 'sentence'),
'qqp': ('question1', 'question2'),
'rte': ('sentence1', 'sentence2'),
'sst2': ('sentence', None),
'stsb': ('sentence1', 'sentence2'),
'wnli': ('sentence1', 'sentence2'),
}
sentence1_key, sentence2_key = task_to_keys[data_name]
# used to preprocess the raw data
def preprocess_function(examples):
args = (
(examples[sentence1_key],) if sentence2_key is None else (
examples[sentence1_key], examples[sentence2_key])
)
result = tokenizer(*args, padding=False, max_length=128, truncation=True)
if 'label' in examples:
result['labels'] = examples['label']
return result
raw_datasets = load_dataset('glue', data_name, cache_dir=cache_dir)
if eval_key == 'val':
for key in list(raw_datasets.keys()):
if 'test' in key:
raw_datasets.pop(key)
column_names = raw_datasets['train'].column_names
processed_datasets = raw_datasets.map(preprocess_function, batched=True, remove_columns=column_names)
if data_name == 'mnli':
if eval_key == 'test':
validation_datasets = {
'test_matched': processed_datasets['validation_matched'],
'test_mismatched': processed_datasets['validation_mismatched']
}
else:
validation_datasets = {
'validation_matched': processed_datasets['validation_matched'],
'validation_mismatched': processed_datasets['validation_mismatched']
}
else:
if eval_key == 'test':
validation_datasets = {
'test': processed_datasets['test']
}
else:
validation_datasets = {
'validation': processed_datasets['validation']
}
return processed_datasets['train'], validation_datasets, None
def prepare_datasets_cifar(model_name: str, data_name: str, tokenizer, cache_dir: str, eval_key: str = 'val'):
train_ds, test_ds = load_dataset(data_name, cache_dir=cache_dir, split=['train', 'test'])
# split up training into training + validation
splits = train_ds.train_test_split(test_size=0.1)
train_ds = splits['train']
val_ds = splits['test']
image_mean, image_std = tokenizer.image_mean, tokenizer.image_std
size = 224
normalize = Normalize(mean=image_mean, std=image_std)
_train_transforms = Compose(
[
RandomResizedCrop(size),
RandomHorizontalFlip(),
ToTensor(),
normalize,
]
)
_val_transforms = Compose(
[
Resize(size),
CenterCrop(size),
ToTensor(),
normalize,
]
)
def train_transform(examples):
examples['pixel_values'] = [_train_transforms(image.convert("RGB")) for image in examples['img']]
return examples
def val_transform(examples):
examples['pixel_values'] = [_val_transforms(image.convert("RGB")) for image in examples['img']]
return examples
train_ds.set_transform(train_transform)
val_ds.set_transform(val_transform)
test_ds.set_transform(val_transform)
return train_ds, val_ds, test_ds
def prepare_datasets_tinyimagenet(model_name: str, data_name: str, tokenizer, cache_dir: str, eval_key: str = 'val'):
train_ds, test_ds = load_dataset('Maysee/tiny-imagenet', cache_dir=cache_dir, split=['train', 'valid'])
# split up training into training + validation
splits = train_ds.train_test_split(test_size=0.1)
train_ds = splits['train']
val_ds = splits['test']
image_mean, image_std = tokenizer.image_mean, tokenizer.image_std
size = 224
normalize = Normalize(mean=image_mean, std=image_std)
_train_transforms = Compose(
[
RandomResizedCrop(size),
RandomHorizontalFlip(),
ToTensor(),
normalize,
]
)
_val_transforms = Compose(
[
Resize(size),
CenterCrop(size),
ToTensor(),
normalize,
]
)
def train_transform(examples):
examples['pixel_values'] = [_train_transforms(image.convert("RGB")) for image in examples['image']]
return examples
def val_transform(examples):
examples['pixel_values'] = [_val_transforms(image.convert("RGB")) for image in examples['image']]
return examples
train_ds.set_transform(train_transform)
val_ds.set_transform(val_transform)
test_ds.set_transform(val_transform)
return train_ds, val_ds, test_ds
def prepare_datasets_cityscapes(model_name: str, data_name: str, tokenizer, cache_dir: str, eval_key: str = 'val'):
num_classes = 19
train_tokenizer = copy(tokenizer)
train_mul = 1
train_tokenizer.size = {'height': int(512 * train_mul), 'width': int(1024 * train_mul)}
eval_tokenizer = copy(tokenizer)
eval_mul = 2
eval_tokenizer.size = {'height': int(512 * eval_mul), 'width': int(1024 * eval_mul)}
train_ds = Cityscapes(
root=cache_dir,
list_path='/list/cityscapes/train.lst',
tokenizer=train_tokenizer,
num_classes=num_classes,
ignore_label=255)
val_ds = Cityscapes(
root=cache_dir,
list_path='/list/cityscapes/val.lst',
tokenizer=eval_tokenizer,
num_classes=num_classes,
ignore_label=255)
test_ds = Cityscapes(
root=cache_dir,
list_path='/list/cityscapes/test.lst',
tokenizer=eval_tokenizer,
num_classes=num_classes,
ignore_label=255)
return train_ds, val_ds, test_ds
def prepare_datasets_kitti(model_name: str, data_name: str, tokenizer, cache_dir: str, eval_key: str = 'val'):
num_classes = 19
train_tokenizer = copy(tokenizer)
train_mul = 1
train_tokenizer.size = {'height': int(375 * train_mul), 'width': int(1242 * train_mul)}
eval_tokenizer = copy(tokenizer)
eval_mul = 1
eval_tokenizer.size = {'height': int(375 * eval_mul), 'width': int(1242 * eval_mul)}
train_ds = Cityscapes(
root=cache_dir,
list_path='/list/kitti/train.lst',
tokenizer=train_tokenizer,
num_classes=num_classes,
ignore_label=255)
val_ds = Cityscapes(
root=cache_dir,
list_path='/list/kitti/val.lst',
tokenizer=eval_tokenizer,
num_classes=num_classes,
ignore_label=255)
return train_ds, val_ds, val_ds
class Cityscapes(data.Dataset):
def __init__(self,
root,
list_path,
tokenizer,
num_classes=19,
ignore_label=255):
super(Cityscapes, self).__init__()
self.tokenizer = tokenizer
self.root = root
self.list_path = list_path
self.num_classes = num_classes
self.img_list = [line.strip().split() for line in open(root + list_path)]
self.files = self.read_files()
self.label_mapping = {-1: ignore_label, 0: ignore_label,
1: ignore_label, 2: ignore_label,
3: ignore_label, 4: ignore_label,
5: ignore_label, 6: ignore_label,
7: 0, 8: 1, 9: ignore_label,
10: ignore_label, 11: 2, 12: 3,
13: 4, 14: ignore_label, 15: ignore_label,
16: ignore_label, 17: 5, 18: ignore_label,
19: 6, 20: 7, 21: 8, 22: 9, 23: 10, 24: 11,
25: 12, 26: 13, 27: 14, 28: 15,
29: ignore_label, 30: ignore_label,
31: 16, 32: 17, 33: 18}
self.target_mode = False
self.image_mode = False
def __len__(self):
return len(self.files)
def read_files(self):
files = []
if 'test' in self.list_path:
for item in self.img_list:
image_path = item
name = os.path.splitext(os.path.basename(image_path[0]))[0]
files.append({
"img": image_path[0],
"name": name,
})
else:
for item in self.img_list:
image_path, label_path = item
name = os.path.splitext(os.path.basename(label_path))[0]
files.append({
"img": image_path,
"label": label_path,
"name": name,
"weight": 1
})
return files
def convert_label(self, label, inverse=False):
temp = label.copy()
if inverse:
for v, k in self.label_mapping.items():
label[temp == k] = v
else:
for k, v in self.label_mapping.items():
label[temp == k] = v
return label
def __getitem__(self, index):
item = self.files[index]
if 'cityscapes' in self.list_path:
folder = 'cityscapes'
else:
folder = 'kitti'
image = cv2.imread(os.path.join(self.root, folder, item["img"]), cv2.IMREAD_COLOR)
if 'test' in self.list_path:
return self.tokenizer(image)
label = cv2.imread(os.path.join(self.root, folder, item["label"]), cv2.IMREAD_GRAYSCALE)
label = self.convert_label(label)
return self.tokenizer(image, label)