-
Notifications
You must be signed in to change notification settings - Fork 11
/
utils.py
382 lines (309 loc) · 11.5 KB
/
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
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
"""
Misc functions, including distributed helpers.
Mostly copy-paste from torchvision references.
"""
import io
import json
import os
import time
from collections import defaultdict, deque
import datetime
import torch
import torch.distributed as dist
from logger import logger
import numpy as np
from itertools import combinations
class SmoothedValue(object):
"""Track a series of values and provide access to smoothed values over a
window or the global series average.
"""
def __init__(self, window_size=20, fmt=None):
if fmt is None:
fmt = "{median:.4f} ({global_avg:.4f})"
self.deque = deque(maxlen=window_size)
self.total = 0.0
self.count = 0
self.fmt = fmt
def update(self, value, n=1):
self.deque.append(value)
self.count += n
self.total += value * n
def synchronize_between_processes(self):
"""
Warning: does not synchronize the deque!
"""
if not is_dist_avail_and_initialized():
return
t = torch.tensor([self.count, self.total], dtype=torch.float64, device='cuda')
dist.barrier()
dist.all_reduce(t)
t = t.tolist()
self.count = int(t[0])
self.total = t[1]
@property
def median(self):
d = torch.tensor(list(self.deque))
return d.median().item()
@property
def avg(self):
d = torch.tensor(list(self.deque), dtype=torch.float32)
return d.mean().item()
@property
def global_avg(self):
return self.total / self.count
@property
def max(self):
return max(self.deque)
@property
def value(self):
return self.deque[-1]
def __str__(self):
return self.fmt.format(
median=self.median,
avg=self.avg,
global_avg=self.global_avg,
max=self.max,
value=self.value)
class MetricLogger(object):
def __init__(self, delimiter="\t"):
self.meters = defaultdict(SmoothedValue)
self.delimiter = delimiter
def update(self, **kwargs):
for k, v in kwargs.items():
if isinstance(v, torch.Tensor):
v = v.item()
assert isinstance(v, (float, int))
self.meters[k].update(v)
def __getattr__(self, attr):
if attr in self.meters:
return self.meters[attr]
if attr in self.__dict__:
return self.__dict__[attr]
raise AttributeError("'{}' object has no attribute '{}'".format(
type(self).__name__, attr))
def __str__(self):
loss_str = []
for name, meter in self.meters.items():
loss_str.append(
"{}: {}".format(name, str(meter))
)
return self.delimiter.join(loss_str)
def synchronize_between_processes(self):
for meter in self.meters.values():
meter.synchronize_between_processes()
def add_meter(self, name, meter):
self.meters[name] = meter
def log_every(self, iterable, print_freq, header=None):
i = 0
if not header:
header = ''
start_time = time.time()
end = time.time()
iter_time = SmoothedValue(fmt='{avg:.4f}')
data_time = SmoothedValue(fmt='{avg:.4f}')
space_fmt = ':' + str(len(str(len(iterable)))) + 'd'
log_msg = [
header,
'[{0' + space_fmt + '}/{1}]',
'eta: {eta}',
'{meters}',
'time: {time}',
'data: {data}'
]
if torch.cuda.is_available():
log_msg.append('max mem: {memory:.0f}')
log_msg = self.delimiter.join(log_msg)
MB = 1024.0 * 1024.0
for obj in iterable:
data_time.update(time.time() - end)
yield obj
iter_time.update(time.time() - end)
if i % print_freq == 0 or i == len(iterable) - 1:
eta_seconds = iter_time.global_avg * (len(iterable) - i)
eta_string = str(datetime.timedelta(seconds=int(eta_seconds)))
if torch.cuda.is_available():
logger.info(log_msg.format(
i, len(iterable), eta=eta_string,
meters=str(self),
time=str(iter_time), data=str(data_time),
memory=torch.cuda.max_memory_allocated() / MB))
else:
logger.info(log_msg.format(
i, len(iterable), eta=eta_string,
meters=str(self),
time=str(iter_time), data=str(data_time)))
i += 1
end = time.time()
total_time = time.time() - start_time
total_time_str = str(datetime.timedelta(seconds=int(total_time)))
logger.info('{} Total time: {} ({:.4f} s / it)'.format(
header, total_time_str, total_time / len(iterable)))
def _load_checkpoint_for_ema(model_ema, checkpoint):
"""
Workaround for ModelEma._load_checkpoint to accept an already-loaded object
"""
mem_file = io.BytesIO()
torch.save(checkpoint, mem_file)
mem_file.seek(0)
model_ema._load_checkpoint(mem_file)
def setup_for_distributed(is_master):
"""
This function disables printing when not in master process
"""
import builtins as __builtin__
builtin_print = __builtin__.print
def print(*args, **kwargs):
force = kwargs.pop('force', False)
if is_master or force:
builtin_print(*args, **kwargs)
__builtin__.print = print
def is_dist_avail_and_initialized():
if not dist.is_available():
return False
if not dist.is_initialized():
return False
return True
def get_world_size():
if not is_dist_avail_and_initialized():
return 1
return dist.get_world_size()
def get_rank():
if not is_dist_avail_and_initialized():
return 0
return dist.get_rank()
def is_main_process():
return get_rank() == 0
def save_on_master(*args, **kwargs):
if is_main_process():
torch.save(*args, **kwargs)
def save_on_master_eval_res(log_stats, output_dir):
if is_main_process():
with open(output_dir, 'a') as f:
f.write(json.dumps(log_stats) + "\n")
def init_distributed_mode(args):
if 'RANK' in os.environ and 'WORLD_SIZE' in os.environ:
args.rank = int(os.environ["RANK"])
args.world_size = int(os.environ['WORLD_SIZE'])
args.gpu = int(os.environ['LOCAL_RANK'])
elif 'SLURM_PROCID' in os.environ:
args.rank = int(os.environ['SLURM_PROCID'])
args.gpu = args.rank % torch.cuda.device_count()
else:
logger.info('Not using distributed mode')
args.distributed = False
return
args.distributed = True
torch.cuda.set_device(args.gpu)
args.dist_backend = 'nccl'
logger.info('| distributed init (rank {}): {}'.format(
args.rank, args.dist_url))
torch.distributed.init_process_group(backend=args.dist_backend, init_method=args.dist_url,
world_size=args.world_size, rank=args.rank)
torch.distributed.barrier()
setup_for_distributed(args.rank == 0)
def paired_stitching(depth=12, kernel_size=2, stride=1):
blk_id = list(range(depth))
i = 0
stitch_cfgs = []
stitch_id = -1
stitching_layers_mappings = []
while i < depth:
ids = blk_id[i:i + kernel_size]
has_new_stitches = False
for j in ids:
for k in ids:
if (j, k) not in stitch_cfgs:
has_new_stitches = True
stitch_cfgs.append((j, k))
stitching_layers_mappings.append(stitch_id + 1)
if has_new_stitches:
stitch_id += 1
i += stride
num_stitches = stitch_id + 1
return stitch_cfgs, stitching_layers_mappings, num_stitches
def unpaired_stitching(front_depth=12, end_depth=24):
num_stitches = front_depth
block_ids = torch.tensor(list(range(front_depth)))
block_ids = block_ids[None, None, :].float()
end_mapping_ids = torch.nn.functional.interpolate(block_ids, end_depth)
end_mapping_ids = end_mapping_ids.squeeze().long().tolist()
front_mapping_ids = block_ids.squeeze().long().tolist()
stitch_cfgs = []
for idx in front_mapping_ids:
for i, e_idx in enumerate(end_mapping_ids):
if idx != e_idx or idx >= i:
continue
else:
stitch_cfgs.append((idx, i))
return stitch_cfgs, end_mapping_ids, num_stitches
def get_stitch_configs(depth=12, kernel_size=2, stride=1, num_models=3, nearest_stitching=True):
'''This function assumes the two model have the same depth, for demonstrating DeiT stitching.
Args:
depth: number of blocks in the model
kernel_size: size of the stitching sliding window
stride: stride of the stitching sliding window
num_models: number of models to be stitched
nearest_stitching: whether to use nearest stitching
'''
stitch_cfgs, layers_mappings, num_stitches = paired_stitching(depth, kernel_size, stride)
model_combinations = []
candidates = list(range(num_models))
for i in range(1, num_models + 1):
model_combinations += list(combinations(candidates, i))
if nearest_stitching:
# remove tiny-base
model_combinations.pop(model_combinations.index((0, 2)))
# remove three model settings
model_combinations.pop(model_combinations.index((0, 1, 2)))
total_configs = []
for comb in model_combinations:
if len(comb) == 1:
total_configs.append({
'comb_id': comb,
'stitch_cfgs': [],
'stitch_layers': []
})
continue
for cfg, layer_mapping_id in zip(stitch_cfgs, layers_mappings):
if len(comb) == 2:
total_configs.append({
'comb_id': comb,
'stitch_cfgs': [cfg],
'stitch_layers': [layer_mapping_id]
})
else:
last_out_id = cfg[1]
for second_cfg, second_layer_mapping_id in zip(stitch_cfgs, layers_mappings):
middle_id, end_id = second_cfg
if middle_id < last_out_id:
continue
total_configs.append({
'comb_id': comb,
'stitch_cfgs': [cfg, second_cfg],
'stitch_layers': [layer_mapping_id, second_layer_mapping_id]
})
return total_configs, num_stitches
def rearrange_activations(activations):
n_channels = activations.shape[-1]
activations = activations.reshape(-1, n_channels)
return activations
def ps_inv(x1, x2):
'''Least-squares solver given feature maps from two anchors.
Source: https://github.com/renyi-ai/drfrankenstein/blob/main/src/comparators/compare_functions/ps_inv.py
'''
x1 = rearrange_activations(x1)
x2 = rearrange_activations(x2)
if not x1.shape[0] == x2.shape[0]:
raise ValueError('Spatial size of compared neurons must match when ' \
'calculating psuedo inverse matrix.')
# Get transformation matrix shape
shape = list(x1.shape)
shape[-1] += 1
# Calculate pseudo inverse
x1_ones = torch.ones(shape)
x1_ones[:, :-1] = x1
A_ones = torch.matmul(torch.linalg.pinv(x1_ones), x2.to(x1_ones.device)).T
# Get weights and bias
w = A_ones[..., :-1]
b = A_ones[..., -1]
return w, b