-
Notifications
You must be signed in to change notification settings - Fork 14
/
main_prune_non_imagenet.py
319 lines (273 loc) · 12.8 KB
/
main_prune_non_imagenet.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
import argparse
import json
import math
import os
import sys
import torch
import torch.nn as nn
import torch.optim as optim
from models.model_base import ModelBase
from tensorboardX import SummaryWriter
from tqdm import tqdm
from models.base.init_utils import weights_init
from utils.common_utils import (get_logger, makedirs, process_config, PresetLRScheduler, str_to_list)
from utils.data_utils import get_dataloader
from utils.network_utils import get_network
from pruner.GraSP import GraSP
def get_args():
parser = argparse.ArgumentParser()
parser.add_argument('--target_ratio', type=float)
parser.add_argument('--iteration', type=int)
parser.add_argument('--reset_to', type=int)
parser.add_argument('--network', type=str, default='vgg')
parser.add_argument('--dataset', type=str, default='cifar10')
parser.add_argument('--depth', type=int, default=19)
parser.add_argument('--pretrain_model', type=str)
parser.add_argument('--log_dir', type=str, default='runs/')
def init_config():
parser = argparse.ArgumentParser()
parser.add_argument('--config', type=str, required=True)
parser.add_argument('--run', type=str, default='')
args = parser.parse_args()
runs = None
if len(args.run) > 0:
runs = args.run
config = process_config(args.config, runs)
return config
def init_logger(config):
makedirs(config.summary_dir)
makedirs(config.checkpoint_dir)
# set logger
path = os.path.dirname(os.path.abspath(__file__))
path_model = os.path.join(path, 'models/base/%s.py' % config.network.lower())
path_main = os.path.join(path, 'main_prune_non_imagenet.py')
path_pruner = os.path.join(path, 'pruner/%s.py' % config.pruner_file)
logger = get_logger('log', logpath=config.summary_dir + '/',
filepath=path_model, package_files=[path_main, path_pruner])
logger.info(dict(config))
writer = SummaryWriter(config.summary_dir)
# sys.stdout = open(os.path.join(config.summary_dir, 'stdout.txt'), 'w+')
# sys.stderr = open(os.path.join(config.summary_dir, 'stderr.txt'), 'w+')
return logger, writer
def print_mask_information(mb, logger):
ratios = mb.get_ratio_at_each_layer()
logger.info('** Mask information of %s. Overall Remaining: %.2f%%' % (mb.get_name(), ratios['ratio']))
count = 0
for k, v in ratios.items():
if k == 'ratio':
continue
logger.info(' (%d) %s: Remaining: %.2f%%' % (count, k, v))
count += 1
def save_state(net, acc, epoch, loss, config, ckpt_path, is_best=False):
print('Saving..')
state = {
'net': net,
'acc': acc,
'epoch': epoch,
'loss': loss,
'args': config
}
if not is_best:
torch.save(state, '%s/pruned_%s_%s%s_%d.t7' % (ckpt_path,
config.dataset,
config.network,
config.depth,
epoch))
else:
torch.save(state, '%s/finetuned_%s_%s%s_best.t7' % (ckpt_path,
config.dataset,
config.network,
config.depth))
def train(net, loader, optimizer, criterion, lr_scheduler, epoch, writer, iteration):
print('\nEpoch: %d' % epoch)
net.train()
train_loss = 0
correct = 0
total = 0
lr_scheduler(optimizer, epoch)
desc = ('[LR=%s] Loss: %.3f | Acc: %.3f%% (%d/%d)' %
(lr_scheduler.get_lr(optimizer), 0, 0, correct, total))
writer.add_scalar('iter_%d/train/lr' % iteration, lr_scheduler.get_lr(optimizer), epoch)
prog_bar = tqdm(enumerate(loader), total=len(loader), desc=desc, leave=True)
for batch_idx, (inputs, targets) in prog_bar:
inputs, targets = inputs.cuda(), targets.cuda()
optimizer.zero_grad()
outputs = net(inputs)
loss = criterion(outputs, targets)
# import pdb; pdb.set_trace()
loss.backward()
optimizer.step()
train_loss += loss.item()
_, predicted = outputs.max(1)
total += targets.size(0)
correct += predicted.eq(targets).sum().item()
desc = ('[LR=%s] Loss: %.3f | Acc: %.3f%% (%d/%d)' %
(lr_scheduler.get_lr(optimizer), train_loss / (batch_idx + 1), 100. * correct / total, correct, total))
prog_bar.set_description(desc, refresh=True)
writer.add_scalar('iter_%d/train/loss' % iteration, train_loss / (batch_idx + 1), epoch)
writer.add_scalar('iter_%d/train/acc' % iteration, 100. * correct / total, epoch)
def test(net, loader, criterion, epoch, writer, iteration):
net.eval()
test_loss = 0
correct = 0
total = 0
desc = ('Loss: %.3f | Acc: %.3f%% (%d/%d)'
% (test_loss / (0 + 1), 0, correct, total))
prog_bar = tqdm(enumerate(loader), total=len(loader), desc=desc, leave=True)
with torch.no_grad():
for batch_idx, (inputs, targets) in prog_bar:
inputs, targets = inputs.cuda(), targets.cuda()
outputs = net(inputs)
loss = criterion(outputs, targets)
test_loss += loss.item()
_, predicted = outputs.max(1)
total += targets.size(0)
correct += predicted.eq(targets).sum().item()
desc = ('Loss: %.3f | Acc: %.3f%% (%d/%d)'
% (test_loss / (batch_idx + 1), 100. * correct / total, correct, total))
prog_bar.set_description(desc, refresh=True)
# Save checkpoint.
acc = 100. * correct / total
writer.add_scalar('iter_%d/test/loss' % iteration, test_loss / (batch_idx + 1), epoch)
writer.add_scalar('iter_%d/test/acc' % iteration, 100. * correct / total, epoch)
return acc
def train_once(mb, net, trainloader, testloader, writer, config, ckpt_path, learning_rate, weight_decay, num_epochs,
iteration, logger):
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(net.parameters(), lr=learning_rate, momentum=0.9, weight_decay=weight_decay)
lr_schedule = {0: learning_rate,
int(num_epochs * 0.5): learning_rate * 0.1,
int(num_epochs * 0.75): learning_rate * 0.01}
lr_scheduler = PresetLRScheduler(lr_schedule)
best_acc = 0
best_epoch = 0
for epoch in range(num_epochs):
train(net, trainloader, optimizer, criterion, lr_scheduler, epoch, writer, iteration=iteration)
test_acc = test(net, testloader, criterion, epoch, writer, iteration)
if test_acc > best_acc:
print('Saving..')
state = {
'net': net,
'acc': test_acc,
'epoch': epoch,
'args': config,
'mask': mb.masks,
'ratio': mb.get_ratio_at_each_layer()
}
path = os.path.join(ckpt_path, 'finetune_%s_%s%s_r%s_it%d_best.pth.tar' % (config.dataset,
config.network,
config.depth,
config.target_ratio,
iteration))
torch.save(state, path)
best_acc = test_acc
best_epoch = epoch
logger.info('Iteration [%d], best acc: %.4f, epoch: %d' %
(iteration, best_acc, best_epoch))
def get_exception_layers(net, exception):
exc = []
idx = 0
for m in net.modules():
if isinstance(m, (nn.Linear, nn.Conv2d)):
if idx in exception:
exc.append(m)
idx += 1
return tuple(exc)
def main(config):
# init logger
classes = {
'cifar10': 10,
'cifar100': 100,
'mnist': 10,
'tiny_imagenet': 200
}
logger, writer = init_logger(config)
# build model
model = get_network(config.network, config.depth, config.dataset, use_bn=config.get('use_bn', True))
mask = None
mb = ModelBase(config.network, config.depth, config.dataset, model)
mb.cuda()
if mask is not None:
mb.register_mask(mask)
print_mask_information(mb, logger)
# preprocessing
# ====================================== get dataloader ======================================
trainloader, testloader = get_dataloader(config.dataset, config.batch_size, 256, 4)
# ====================================== fetch configs ======================================
ckpt_path = config.checkpoint_dir
num_iterations = config.iterations
target_ratio = config.target_ratio
normalize = config.normalize
# ====================================== fetch exception ======================================
exception = get_exception_layers(mb.model, str_to_list(config.exception, ',', int))
logger.info('Exception: ')
for idx, m in enumerate(exception):
logger.info(' (%d) %s' % (idx, m))
# ====================================== fetch training schemes ======================================
ratio = 1 - (1 - target_ratio) ** (1.0 / num_iterations)
learning_rates = str_to_list(config.learning_rate, ',', float)
weight_decays = str_to_list(config.weight_decay, ',', float)
training_epochs = str_to_list(config.epoch, ',', int)
logger.info('Normalize: %s, Total iteration: %d, Target ratio: %.2f, Iter ratio %.4f.' %
(normalize, num_iterations, target_ratio, ratio))
logger.info('Basic Settings: ')
for idx in range(len(learning_rates)):
logger.info(' %d: LR: %.5f, WD: %.5f, Epochs: %d' % (idx,
learning_rates[idx],
weight_decays[idx],
training_epochs[idx]))
# ====================================== start pruning ======================================
iteration = 0
for _ in range(1):
logger.info('** Target ratio: %.4f, iter ratio: %.4f, iteration: %d/%d.' % (target_ratio,
ratio,
1,
num_iterations))
mb.model.apply(weights_init)
print("=> Applying weight initialization(%s)." % config.get('init_method', 'kaiming'))
print("Iteration of: %d/%d" % (iteration, num_iterations))
masks = GraSP(mb.model, ratio, trainloader, 'cuda',
num_classes=classes[config.dataset],
samples_per_class=config.samples_per_class,
num_iters=config.get('num_iters', 1))
iteration = 0
print('=> Using GraSP')
# ========== register mask ==================
mb.register_mask(masks)
# ========== save pruned network ============
logger.info('Saving..')
state = {
'net': mb.model,
'acc': -1,
'epoch': -1,
'args': config,
'mask': mb.masks,
'ratio': mb.get_ratio_at_each_layer()
}
path = os.path.join(ckpt_path, 'prune_%s_%s%s_r%s_it%d.pth.tar' % (config.dataset,
config.network,
config.depth,
config.target_ratio,
iteration))
torch.save(state, path)
# ========== print pruning details ============
logger.info('**[%d] Mask and training setting: ' % iteration)
print_mask_information(mb, logger)
logger.info(' LR: %.5f, WD: %.5f, Epochs: %d' %
(learning_rates[iteration], weight_decays[iteration], training_epochs[iteration]))
# ========== finetuning =======================
train_once(mb=mb,
net=mb.model,
trainloader=trainloader,
testloader=testloader,
writer=writer,
config=config,
ckpt_path=ckpt_path,
learning_rate=learning_rates[iteration],
weight_decay=weight_decays[iteration],
num_epochs=training_epochs[iteration],
iteration=iteration,
logger=logger)
if __name__ == '__main__':
config = init_config()
main(config)