/
train_ss_mn.py
152 lines (124 loc) · 5.9 KB
/
train_ss_mn.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
import os, sys, time
import shutil
import yaml
import argparse
import chainer
from chainer import training
from chainer.training import extension
from chainer.training import extensions
base = os.path.dirname(os.path.abspath(__file__))
sys.path.append(os.path.join(base, '../../'))
import source.yaml_utils as yaml_utils
import multiprocessing
from extentions import validation_loss_and_acc, adversarial_validation_loss_and_acc
import json
import functools
import warnings
def create_result_dir(result_dir, config_path, config):
if not os.path.exists(result_dir):
os.makedirs(result_dir)
def copy_to_result_dir(fn, result_dir):
bfn = os.path.basename(fn)
shutil.copy(fn, '{}/{}'.format(result_dir, bfn))
open(os.path.join(result_dir, os.path.basename(config_path)), 'w').write(
yaml.dump(config, default_flow_style=False))
copy_to_result_dir(
config['models']['classifier']['fn'], result_dir)
copy_to_result_dir(
config['dataset']['dataset_fn'], result_dir)
copy_to_result_dir(
config['updater']['fn'], result_dir)
def load_models(config):
cls_conf = config['models']['classifier']
cls = yaml_utils.load_model(cls_conf['fn'], cls_conf['name'], cls_conf['args'])
return cls
def load_dataset_eval(config):
dataset = yaml_utils.load_module(config.dataset_eval['dataset_fn'],
config.dataset_eval['dataset_name'])
return dataset(**config.dataset_eval['args'])
def make_optimizer(model, alpha=0.001, beta1=0.9, beta2=0.999):
optimizer = chainer.optimizers.Adam(alpha=alpha, beta1=beta1, beta2=beta2)
optimizer.setup(model)
return optimizer
def main(config, args):
out = args.results_dir
create_result_dir(out, args.config_path, config)
config = yaml_utils.Config(config)
chainer.cuda.get_device_from_id(0).use()
print("init")
classifier = load_models(config)
classifier.to_gpu()
# Optimizer
opt = make_optimizer(
classifier, alpha=config.adam['alpha'], beta1=config.adam['beta1'], beta2=config.adam['beta2'])
dataset = yaml_utils.load_dataset(config)
dataset_l = dataset.dataset_l
dataset_ul = dataset.dataset_ul
# Iterator
multiprocessing.set_start_method('forkserver')
iterator_l = chainer.iterators.SerialIterator(dataset_l, config.batchsize)
iterator_ul = chainer.iterators.SerialIterator(dataset_ul, config.batchsize_ul)
iterators = {'main': iterator_l, 'unlabeled': iterator_ul}
kwargs = config.updater['args'] if 'args' in config.updater else {}
kwargs.update({
'classifier': classifier,
'iterator': iterators,
'optimizer': opt,
})
updater = yaml_utils.load_updater_class(config)
updater = updater(**kwargs)
trainer = training.Trainer(updater, (config.iteration, 'iteration'), out=out)
report_keys = ["loss_total", "loss_l", "loss_adv", "loss_ul", "loss_ul_separated",
"loss_vadv", "val_phi", "val_phi_0", "val_phi_1", "val_phi_2", "val_loss", "adv_val_loss", "acc",
"val_acc", "adv_val_acc"]
# Set up logging
trainer.extend(extensions.snapshot(), trigger=(config.snapshot_interval, 'iteration'))
trainer.extend(extensions.snapshot_object(
classifier, classifier.__class__.__name__ + '_{.updater.iteration}.npz'),
trigger=(config.snapshot_interval, 'iteration'))
trainer.extend(extensions.LogReport(keys=report_keys,
trigger=(config.display_interval, 'iteration')))
trainer.extend(extensions.PrintReport(report_keys), trigger=(config.display_interval, 'iteration'))
# Eval dataset and iterator
dataset_eval = load_dataset_eval(config)
eval_iter = chainer.iterators.SerialIterator(dataset_eval, 100, shuffle=False)
trainer.extend(validation_loss_and_acc(classifier, eval_iter, loss_type=config.loss_type, n=5000),
trigger=(config.evaluation_interval, 'iteration'),
priority=extension.PRIORITY_WRITER)
trainer.extend(adversarial_validation_loss_and_acc(classifier, eval_iter,
steps=config.steps, gamma=config.gamma, alpha=config.alpha,
loss_type=config.loss_type, c_type=config.c_type,
clip_x=config.clip_x, n=5000),
trigger=(config.evaluation_interval, 'iteration'),
priority=extension.PRIORITY_WRITER)
trainer.extend(extensions.ProgressBar(update_interval=config.display_interval))
trainer.extend(extensions.LinearShift('alpha', (config.adam['alpha'], 0.),
(config.iteration_decay_start, config.iteration), opt))
if args.snapshot:
print("Resume training with snapshot:{}".format(args.snapshot))
chainer.serializers.load_npz(args.snapshot, trainer)
print("start training")
trainer.run()
def parse_args():
parser = argparse.ArgumentParser()
parser.add_argument('--config_path', type=str, default='configs/base.yml', help='path to config file')
parser.add_argument('--results_dir', type=str, default='./results')
parser.add_argument('--snapshot', type=str, default='',
help='path to the snapshot')
parser.add_argument('-a', '--attrs', nargs='*', default=())
parser.add_argument('-w', '--warning', action='store_true')
args = parser.parse_args()
config = yaml.load(open(args.config_path))
for attr in args.attrs:
module, new_value = attr.split('=')
keys = module.split('.')
target = functools.reduce(dict.__getitem__, keys[:-1], config)
target[keys[-1]] = yaml.load(new_value)
print(config)
return config, args
if __name__ == '__main__':
config, args = parse_args()
if not args.warning:
# Ignore warnings
warnings.simplefilter('ignore')
main(config, args)