-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathtrain.py
230 lines (194 loc) · 10.5 KB
/
train.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
import os
import datetime
from pickle import FALSE
import numpy as np
import torch
import torch.backends.cudnn as cudnn
import torch.distributed as dist
import torch.optim as optim
from torch.utils.data import DataLoader
from nets.AECIF_Net import AECIF_Net
from nets.hrnet_training import (get_lr_scheduler, set_optimizer_lr,
weights_init)
from utils.callbacks_multi import LossHistory, EvalCallback
from utils.dataloader_multi import SegmentationDataset, seg_dataset_collate
from utils.utils import download_weights, show_config
from utils.utils_fit_DWA import fit_one_epoch
if __name__ == "__main__":
Cuda = True #GPU/CPU
distributed = False
sync_bn = False #sync_bn
fp16 = True
num_classes = [7,2] #num_classes
backbone = "hrnetv2_w48" #hrnetv2_w18/hrnetv2_w32/hrnetv2_w48
pretrained = False #pretrained
model_path = "model_data/best_checkpoint_86.76_PSA_s.pth"
input_shape = [520, 520]
Init_Epoch = 0
Freeze_Epoch = 15
Freeze_batch_size = 8
UnFreeze_Epoch = 150
Unfreeze_batch_size = 8
Freeze_Train = False
Init_lr = 5e-4
Min_lr = Init_lr * 0.01
optimizer_type = "adam" #adam/sgd
momentum = 0.9
weight_decay = 0
lr_decay_type = 'cos' #step/cos/exp
save_period = 10
save_dir = 'logs'
eval_flag = True
eval_period = 10
VOCdevkit_path = 'VOCdevkit'
dice_loss = False
focal_loss = False
cls_weights = np.array([[1, 1, 1, 1, 1, 1, 1], [1, 1]], object)
num_workers = 4
ngpus_per_node = torch.cuda.device_count()
if distributed:
dist.init_process_group(backend="nccl")
local_rank = int(os.environ["LOCAL_RANK"])
rank = int(os.environ["RANK"])
device = torch.device("cuda", local_rank)
if local_rank == 0:
print(f"[{os.getpid()}] (rank = {rank}, local_rank = {local_rank}) training...")
print("Gpu Device Count : ", ngpus_per_node)
else:
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
local_rank = 0
if pretrained:
if distributed:
if local_rank == 0:
download_weights(backbone)
dist.barrier()
else:
download_weights(backbone)
model = AECIF_Net(num_classes=num_classes, backbone=backbone, pretrained=pretrained)
if not pretrained:
weights_init(model)
if model_path != '':
if local_rank == 0:
print('Load weights {}.'.format(model_path))
model_dict = model.state_dict()
checkpoint = torch.load(model_path, map_location = device)
pretrained_dict = checkpoint['state_dict']
load_key, no_load_key, temp_dict = [], [], {}
for k, v in pretrained_dict.items():
name = k[7:]
if name in model_dict.keys() and np.shape(model_dict[name]) == np.shape(v):
temp_dict[name] = v
load_key.append(name)
else:
no_load_key.append(name)
model_dict.update(temp_dict)
if local_rank == 0:
print("\nSuccessful Load Key:", str(load_key)[:500], "……\nSuccessful Load Key Num:", len(load_key))
print("\nFail To Load Key:", str(no_load_key)[:500], "……\nFail To Load Key num:", len(no_load_key))
if local_rank == 0:
time_str = datetime.datetime.strftime(datetime.datetime.now(),'%Y_%m_%d_%H_%M_%S')
log_dir = os.path.join(save_dir, "loss_" + str(time_str))
loss_history = LossHistory(log_dir, model, input_shape=input_shape)
else:
loss_history = None
if fp16:
from torch.cuda.amp import GradScaler as GradScaler
scaler = GradScaler()
else:
scaler = None
model_train = model.train()
if sync_bn and ngpus_per_node > 1 and distributed:
model_train = torch.nn.SyncBatchNorm.convert_sync_batchnorm(model_train)
elif sync_bn:
print("Sync_bn is not support in one gpu or not distributed.")
if Cuda:
if distributed:
model_train = model_train.cuda(local_rank)
model_train = torch.nn.parallel.DistributedDataParallel(model_train, device_ids=[local_rank], find_unused_parameters=True)
else:
model_train = torch.nn.DataParallel(model)
cudnn.benchmark = True
model_train = model_train.cuda()
with open(os.path.join(VOCdevkit_path, "VOC2007/ImageSets/Segmentation/train.txt"),"r") as f:
train_lines = f.readlines()
with open(os.path.join(VOCdevkit_path, "VOC2007/ImageSets/Segmentation/test.txt"),"r") as f:
val_lines = f.readlines()
num_train = len(train_lines)
num_val = len(val_lines)
if local_rank == 0:
show_config(
num_classes = num_classes, backbone = backbone, model_path = model_path, input_shape = input_shape, \
Init_Epoch = Init_Epoch, Freeze_Epoch = Freeze_Epoch, UnFreeze_Epoch = UnFreeze_Epoch, Freeze_batch_size = Freeze_batch_size, Unfreeze_batch_size = Unfreeze_batch_size, Freeze_Train = Freeze_Train, \
Init_lr = Init_lr, Min_lr = Min_lr, optimizer_type = optimizer_type, momentum = momentum, lr_decay_type = lr_decay_type, \
save_period = save_period, save_dir = save_dir, num_workers = num_workers, num_train = num_train, num_val = num_val
)
wanted_step = 1.5e4 if optimizer_type == "sgd" else 0.5e4
total_step = num_train // Unfreeze_batch_size * UnFreeze_Epoch
if total_step <= wanted_step:
wanted_epoch = wanted_step // (num_train // Unfreeze_batch_size) + 1
if True:
UnFreeze_flag = False
if Freeze_Train:
for param in model.backbone.parameters():
param.requires_grad = False
batch_size = Freeze_batch_size if Freeze_Train else Unfreeze_batch_size
nbs = 16
lr_limit_max = 5e-4 if optimizer_type == 'adam' else 1e-1
lr_limit_min = 3e-4 if optimizer_type == 'adam' else 5e-4
Init_lr_fit = min(max(batch_size / nbs * Init_lr, lr_limit_min), lr_limit_max)
Min_lr_fit = min(max(batch_size / nbs * Min_lr, lr_limit_min * 1e-2), lr_limit_max * 1e-2)
optimizer = optim.Adam(model.parameters(), Init_lr_fit, betas = (momentum, 0.999), weight_decay = weight_decay)
lr_scheduler_func = get_lr_scheduler(lr_decay_type, Init_lr_fit, Min_lr_fit, UnFreeze_Epoch)
epoch_step = num_train // batch_size
epoch_step_val = num_val // batch_size
if epoch_step == 0 or epoch_step_val == 0:
raise ValueError("Please expand your dataset, the dataset is too small to train.")
train_dataset = SegmentationDataset(train_lines, input_shape, num_classes, True, VOCdevkit_path)
val_dataset = SegmentationDataset(val_lines, input_shape, num_classes, False, VOCdevkit_path)
if distributed:
train_sampler = torch.utils.data.distributed.DistributedSampler(train_dataset, shuffle=True,)
val_sampler = torch.utils.data.distributed.DistributedSampler(val_dataset, shuffle=False,)
batch_size = batch_size // ngpus_per_node
shuffle = False
else:
train_sampler = None
val_sampler = None
shuffle = True
gen = DataLoader(train_dataset, shuffle = shuffle, batch_size = batch_size, num_workers = num_workers, pin_memory=True,
drop_last = True, collate_fn = seg_dataset_collate, sampler=train_sampler)
gen_val = DataLoader(val_dataset , shuffle = shuffle, batch_size = batch_size, num_workers = num_workers, pin_memory=True,
drop_last = True, collate_fn = seg_dataset_collate, sampler=val_sampler)
if local_rank == 0:
eval_callback = EvalCallback(model, input_shape, num_classes, val_lines, VOCdevkit_path, log_dir, Cuda, \
eval_flag=eval_flag, period=eval_period)
else:
eval_callback = None
for epoch in range(Init_Epoch, UnFreeze_Epoch):
if epoch >= Freeze_Epoch and not UnFreeze_flag and Freeze_Train:
batch_size = Unfreeze_batch_size
nbs = 16
lr_limit_max = 5e-4 if optimizer_type == 'adam' else 1e-1
lr_limit_min = 3e-4 if optimizer_type == 'adam' else 5e-4
Init_lr_fit = min(max(batch_size / nbs * Init_lr, lr_limit_min), lr_limit_max)
Min_lr_fit = min(max(batch_size / nbs * Min_lr, lr_limit_min * 1e-2), lr_limit_max * 1e-2)
lr_scheduler_func = get_lr_scheduler(lr_decay_type, Init_lr_fit, Min_lr_fit, UnFreeze_Epoch)
for param in model.backbone.parameters():
param.requires_grad = True
epoch_step = num_train // batch_size
epoch_step_val = num_val // batch_size
if epoch_step == 0 or epoch_step_val == 0:
raise ValueError("Please expand your dataset, the dataset is too small to train.")
gen = DataLoader(train_dataset, shuffle = shuffle, batch_size = batch_size, num_workers = num_workers, pin_memory=True,
drop_last = True, collate_fn = seg_dataset_collate, sampler=train_sampler)
gen_val = DataLoader(val_dataset , shuffle = shuffle, batch_size = batch_size, num_workers = num_workers, pin_memory=True,
drop_last = True, collate_fn = seg_dataset_collate, sampler=val_sampler)
UnFreeze_flag = True
if distributed:
train_sampler.set_epoch(epoch)
set_optimizer_lr(optimizer, lr_scheduler_func, epoch)
fit_one_epoch(model_train, model, loss_history, eval_callback, optimizer, epoch, epoch_step, epoch_step_val, gen, gen_val, UnFreeze_Epoch, Cuda, \
dice_loss, focal_loss, cls_weights, num_classes, fp16, scaler, save_period, save_dir, local_rank)
if distributed:
dist.barrier()
if local_rank == 0:
loss_history.writer.close()