-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmain_ft_mp.py
executable file
·357 lines (319 loc) · 15.4 KB
/
main_ft_mp.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
from data_process.datasets import UcfFineTune, UcfFineTuneLMDB, Kin400FTOfflineLMDB
from data_process.preprocess_data import get_transforms
import os
import torch
from torch import nn
from torch import optim
from models.model import generate_model
from opts import parse_opts
from loss import NTXent
from utils import Logger
import numpy as np
import random
import builtins
from utils import get_dataloader
from utils import AverageMeter, calculate_accuracy
import time
import torch.distributed as dist
from scheduler.cosine_anneal import CosineAnnealingWarmupRestarts
def reduce_mean(tensor, world_size):
rt = tensor.clone()
dist.all_reduce(rt, op=dist.ReduceOp.SUM)
rt /= world_size
return rt
def main(opts):
torch.manual_seed(opts.manual_seed)
np.random.seed(opts.manual_seed)
random.seed(opts.manual_seed)
if torch.cuda.is_available():
if opts.local_rank != -1:
opts.cuda = True
opts.world_size = int(os.environ["WORLD_SIZE"])
opts.distributed = True
opts.nprocs = torch.cuda.device_count()
main_worker(opts.local_rank, opts.nprocs, opts)
else:
opts.distributed = False
opts.cuda = True
main_worker('cuda:0', opts.nprocs, opts)
else:
print("-------No CUDA Training!--------")
opts.distributed = False
opts.cuda = False
main_worker("cpu", opts.nprocs, opts)
def main_worker(local_rank, ngpus_per_node, opts):
opts.device = local_rank if opts.cuda else 'cpu'
if opts.distributed:
# suppress printing if not master
if local_rank != 0:
def print_pass(*opts):
pass
builtins.print = print_pass
opts.rank = local_rank
dist.init_process_group(backend=opts.dist_backend,
init_method=opts.dist_url,
world_size=opts.world_size,
rank=opts.rank)
# build log
log_path = os.path.join(opts.result_path, opts.dataset, opts.task)
if not os.path.exists(log_path) and local_rank == 0:
os.makedirs(log_path)
else:
log_path = os.path.join(opts.result_path, opts.dataset, opts.task)
if not os.path.exists(log_path):
os.makedirs(log_path)
# print opts
print(opts)
opts.arch = '{}-{}'.format(opts.model_name, opts.model_depth)
# define transforms and dataloader
if opts.task in ['ft_fc', 'ft_all', 'scratch', 'resume']:
# Tran transform
train_transform = get_transforms(mode=opts.transform_mode, opts=opts)
print("Preprocessing train data ...")
train_data = globals()['{}'.format(opts.dataset)](data_type='train',
opts=opts,
split=opts.split,
sp_transform=train_transform)
len_train_data = len(train_data)
print("Length of training data = ", len_train_data)
# Train dataloader
train_dataloader, train_sampler = get_dataloader(train_data, opts=opts, data_type='train')
# Val transform
val_transform = get_transforms(mode='{}_val'.format(opts.transform_mode), opts=opts)
print("Preprocessing validation data ...")
# Val dataloader
val_data = globals()['{}'.format(opts.dataset)](data_type='val',
opts=opts,
split=opts.split,
sp_transform=val_transform)
len_val_data = len(val_data)
print("Length of validation data = ", len_val_data)
val_dataloader, val_sampler = get_dataloader(val_data, opts=opts, data_type='val')
# Load the model
print("Loading model... ", opts.model_name, opts.model_depth)
model, parameters = generate_model(opts)
criterion_cls = nn.CrossEntropyLoss().cuda(opts.device)
criterion_ctr = NTXent.NTXentLoss(device=opts.device,
batch_size=opts.batch_size,
temperature=opts.temperature,
use_cosine_similarity=True).cuda(opts.device)
criterion = [criterion_cls, criterion_ctr]
if opts.task == 'resume':
begin_epoch = int(opts.resume_md_path.split('/')[-1].split('_')[1])
train_logger = Logger(os.path.join(log_path, '{}_train_clip{}model{}{}.log'.format(
opts.dataset, opts.sample_duration, opts.model_name, opts.model_depth)),
['epoch', 'loss', 'acc', 'lr'], overlay=False)
val_logger = Logger(os.path.join(log_path, '{}_val_clip{}model{}{}.log'.format(
opts.dataset, opts.sample_duration, opts.model_name, opts.model_depth)),
['epoch', 'loss', 'acc'], overlay=False)
else:
begin_epoch = 1
train_logger = Logger(os.path.join(log_path, '{}_train_clip{}model{}{}.log'.format(
opts.dataset, opts.sample_duration, opts.model_name, opts.model_depth)),
['epoch', 'loss', 'acc', 'lr'], overlay=True)
val_logger = Logger(os.path.join(log_path, '{}_val_clip{}model{}{}.log'.format(
opts.dataset, opts.sample_duration, opts.model_name, opts.model_depth)),
['epoch', 'loss', 'acc'], overlay=True)
# build optimizer
if opts.optimizer == 'sgd':
optimizer = optim.SGD(parameters,
lr=opts.learning_rate,
momentum=opts.momentum,
weight_decay=opts.weight_decay)
elif opts.optimizer == 'adamw':
optimizer = optim.AdamW(parameters,
lr=opts.learning_rate,
betas=(0.9, 0.99),
weight_decay=opts.weight_decay)
elif opts.optimizer == 'adam':
optimizer = optim.Adam(parameters,
lr=opts.learning_rate,
weight_decay=opts.weight_decay)
if opts.task == 'resume':
optimizer.load_state_dict(torch.load(opts.resume_md_path)['optimizer'])
# build learning rate strategy
scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer, 'min', patience=opts.lr_patience)
# scheduler = optim.lr_scheduler.StepLR(optimizer, step_size=60, gamma=0.1)
# scheduler = optim.lr_scheduler.CosineAnnealingLR(optimizer, opts.n_epochs, eta_min=opts.lr_decay*opts.learning_rate)
# scheduler = CosineAnnealingWarmupRestarts(
# optimizer, first_cycle_steps=300, cycle_mult=1.0, max_lr=0.03, min_lr=0.00001, warmup_steps=10, gamma=0.5)
# scheduler = CosineAnnealingWarmupRestarts(optimizer,
# first_cycle_steps=opts.n_epochs,
# cycle_mult=1.0,
# max_lr=opts.learning_rate,
# min_lr=0.00001,
# warmup_steps=0.1*opts.n_epochs,
# gamma=0.5)
torch.backends.cudnn.benchmark = True
# Training and Validation
if opts.task in ['ft_fc', 'ft_all', 'scratch', 'resume']:
for epoch in range(begin_epoch, opts.n_epochs + 1):
print('Start to fine-tune')
print('Start training epoch {}'.format(epoch))
if opts.distributed:
train_sampler.set_epoch(epoch)
train(epoch, train_dataloader, model, criterion, optimizer, opts, train_logger, len_train_data)
print('Start validating epoch {}'.format(epoch))
validation(epoch, val_dataloader, model, criterion, optimizer, opts, val_logger, len_val_data, scheduler)
# scheduler.step()
def train(epoch, train_dataloader, model, criterion, optimizer, opts, train_logger, len_train_data):
model.train()
batch_time = AverageMeter()
data_time = AverageMeter()
losses = AverageMeter()
accuracies = AverageMeter()
end_time = time.time()
# CrossEntropy
criterion_cls = criterion[0]
train_prefetcher = data_prefetcher(train_dataloader, opts)
inputs, targets = train_prefetcher.next()
i = 0
while inputs is not None:
i += 1
data_time.update(time.time() - end_time)
assert opts.task in ['scratch', 'r_cls', 'ft_fc', 'ft_all']
# inputs = inputs.to(opts.device, non_blocking=True)
# targets = inputs.to(opts.device, non_blocking=True)
if opts.task in ['']:
outputs = model(inputs, None, 'cls_nofc')
elif opts.task in ['ft_fc', 'ft_all', 'r_cls', 'scratch']:
outputs = model(inputs, o_type=opts.task)
loss = criterion_cls(outputs, targets)
acc = calculate_accuracy(outputs, targets)
# dist.barrier()
reduced_loss = reduce_mean(loss, opts.world_size)
# reduced_acc = reduce_mean(acc, opts.world_size)
losses.update(reduced_loss.item(), inputs.size(0))
accuracies.update(acc, inputs.size(0))
optimizer.zero_grad()
loss.backward()
optimizer.step()
batch_time.update(time.time() - end_time)
end_time = time.time()
print('Epoch: [{0}][{1}/{2}]\t'
'Time {batch_time.val:.3f} ({batch_time.avg:.3f})\t'
'Data {data_time.val:.3f} ({data_time.avg:.3f})\t'
'Loss {loss.val:.4f} ({loss.avg:.4f})\t'
'Acc {acc.val:.3f} ({acc.avg:.3f})\t'
'Lr {lr:.6f}\t'
"Left {left:.1f}d".format(
epoch,
i,
int(len_train_data / opts.batch_size),
batch_time=batch_time,
data_time=data_time,
loss=losses,
acc=accuracies,
lr=optimizer.param_groups[-1]['lr'],
left=(batch_time.avg * ((opts.n_epochs - epoch) * int(len_train_data / opts.batch_size) +
int(len_train_data / opts.batch_size) - i)) / 3600 / 24))
inputs, targets = train_prefetcher.next()
if opts.rank == 0 and opts.local_rank == 0:
train_logger.log({
'epoch': epoch,
'loss': losses.avg,
'acc': accuracies.avg,
'lr': float('{:.5f}'.format(optimizer.param_groups[-1]['lr']))
})
def validation(epoch, val_dataloader, model, criterion, optimizer, opts, val_logger, len_val_data, scheduler):
def step(inputs, targets, i):
end_time = time.time()
data_time.update(time.time() - end_time)
assert opts.task in ['scratch', 'r_cls', 'ft_fc', 'ft_all']
if opts.task in ['']:
outputs = (inputs, None, 'cls_nofc')
elif opts.task in ['ft_fc', 'ft_all', 'r_cls', 'scratch']:
outputs = model(inputs, o_type=opts.task)
loss = criterion(outputs, targets)
acc = calculate_accuracy(outputs, targets)
losses.update(loss.item(), inputs.size(0))
accuracies.update(acc, inputs.size(0))
batch_time.update(time.time() - end_time)
end_time = time.time()
print('Val_Epoch: [{0}][{1}/{2}]\t'
'Time {batch_time.val:.3f} ({batch_time.avg:.3f})\t'
'Data {data_time.val:.3f} ({data_time.avg:.3f})\t'
'Loss {loss.val:.4f} ({loss.avg:.4f})\t'
'Acc {acc.val:.3f} ({acc.avg:.3f})'.format(epoch,
i + 1,
int(len_val_data / opts.batch_size),
batch_time=batch_time,
data_time=data_time,
loss=losses,
acc=accuracies))
batch_time = AverageMeter()
data_time = AverageMeter()
losses = AverageMeter()
accuracies = AverageMeter()
criterion = criterion[0]
# change to evaluation mode
model.eval()
with torch.no_grad():
val_prefetcher = data_prefetcher(val_dataloader, opts)
inputs, targets = val_prefetcher.next()
i = 0
while inputs is not None:
step(inputs, targets, i)
i += 1
inputs, targets = val_prefetcher.next()
if opts.rank == 0 and opts.local_rank == 0:
scheduler.step(losses.avg)
val_logger.log({'epoch': epoch, 'loss': losses.avg, 'acc': accuracies.avg})
accuracy_val = accuracies.avg
if accuracy_val > list(opts.highest_val.values())[0]:
old_key = list(opts.highest_val.keys())[0]
file_path = os.path.join(opts.result_path, opts.dataset, opts.task, old_key)
if os.path.exists(file_path):
os.remove(file_path)
opts.highest_val.pop(old_key)
opts.highest_val['save_{}_max.pth'.format(epoch)] = accuracy_val
save_file_path = os.path.join(opts.result_path,
opts.dataset,
opts.task,
'save_{}_max.pth'.format(epoch))
states = {'epoch': epoch + 1,
'arch': opts.arch,
'state_dict': model.state_dict(),
'optimizer': optimizer.state_dict()}
torch.save(states, save_file_path)
class data_prefetcher():
def __init__(self, loader, opts):
self.loader = iter(loader)
self.opts = opts
self.stream = torch.cuda.Stream()
# self.mean = torch.tensor([0.485 * 255, 0.456 * 255, 0.406 * 255]).cuda().view(1, 3, 1, 1)
# self.std = torch.tensor([0.229 * 255, 0.224 * 255, 0.225 * 255]).cuda().view(1, 3, 1, 1)
# With Amp, it isn't necessary to manually convert data to half.
# if args.fp16:
# self.mean = self.mean.half()
# self.std = self.std.half()
self.preload()
def preload(self):
try:
self.next_input, self.next_target = next(self.loader)
except StopIteration:
self.next_input = None
self.next_target = None
return
with torch.cuda.stream(self.stream):
self.next_input = self.next_input.to(self.opts.device, non_blocking=True)
self.next_target = self.next_target.to(self.opts.device, non_blocking=True)
# With Amp, it isn't necessary to manually convert data to half.
# if args.fp16:
# self.next_input = self.next_input.half()
# else:
# self.next_input = self.next_input.float()
# self.next_input = self.next_input.sub_(self.mean).div_(self.std)
def next(self):
torch.cuda.current_stream().wait_stream(self.stream)
input = self.next_input
target = self.next_target
if input is not None:
input.record_stream(torch.cuda.current_stream())
if target is not None:
target.record_stream(torch.cuda.current_stream())
self.preload()
return input, target
if __name__ == "__main__":
opts = parse_opts()
main(opts)