-
Notifications
You must be signed in to change notification settings - Fork 1
/
train_ALformer.py
234 lines (188 loc) · 8.92 KB
/
train_ALformer.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
import os
from config import Config
opt = Config('train_ALformer.yml')
gpus = ','.join([str(i) for i in opt.GPU])
os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
os.environ["CUDA_VISIBLE_DEVICES"] = gpus
import torch
torch.backends.cudnn.benchmark = True
from thop import profile
from SSIM import SSIM
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
from torch.utils.data import DataLoader
import random
import time
import numpy as np
import utils
from data_RGB import get_training_data, get_validation_data
#from MSGN1G import MSGN
#from STAN import STAN #STRN_64_10
from ALformer import ALformer
#from MPRNet import MPRNet
import losses
#from warmup_scheduler import GradualWarmupScheduler
from tqdm import tqdm
from pdb import set_trace as stx
file1 = 'ALformer_PAR.txt'
file2 = 'SALformer_PSNR.txt'
file3 = 'ALformer_LOSS.txt'
######### Set Seeds ###########
random.seed(1234)
np.random.seed(1234)
torch.manual_seed(1234)
torch.cuda.manual_seed_all(1234)
start_epoch = 1
mode = opt.MODEL.MODE
session = opt.MODEL.SESSION
result_dir = os.path.join(opt.TRAINING.SAVE_DIR, mode, 'results', session)
model_dir = os.path.join(opt.TRAINING.SAVE_DIR, mode, 'models', session)
utils.mkdir(result_dir)
utils.mkdir(model_dir)
train_dir = opt.TRAINING.TRAIN_DIR
val_dir = opt.TRAINING.VAL_DIR
######### Model ###########
#model_restoration = MPRNet()
model_restoration = ALformer()
model_restoration.cuda()
device_ids = [i for i in range(torch.cuda.device_count())]
if torch.cuda.device_count() > 1:
print("\n\nLet's use", torch.cuda.device_count(), "GPUs!\n\n")
initial_lr = opt.OPTIM.LR_INITIAL
optimizer = optim.Adam(model_restoration.parameters(), lr=initial_lr, betas=(0.9, 0.999),eps=1e-8)
######### Scheduler ###########
#warmup_epochs = 3
#scheduler_cosine = optim.lr_scheduler.CosineAnnealingLR(optimizer, opt.OPTIM.NUM_EPOCHS-warmup_epochs, eta_min=opt.OPTIM.LR_MIN)
#scheduler = GradualWarmupScheduler(optimizer, multiplier=1, total_epoch=warmup_epochs, after_scheduler=scheduler_cosine)
#scheduler.step()
############### me
#lambda1=lambda epoch:0.1**(epoch//20)
#lambda2=lambda epoch:0.95** epoch
#scheduler = optim.lr_scheduler.LambdaLR(optimizer,lr_lambda=lambda2,last_epoch=-1)
scheduler = optim.lr_scheduler.StepLR(step_size=80, gamma=0.8, optimizer=optimizer)####step_size epoch
#scheduler = optim.lr_scheduler.StepLR(step_size=40, gamma=0.8, optimizer=optimizer)####step_size epoch
scheduler.step()
######### Resume ###########
if opt.TRAINING.RESUME:
path_chk_rest = utils.get_last_path(model_dir, '_latest.pth')
utils.load_checkpoint(model_restoration,path_chk_rest)
start_epoch = utils.load_start_epoch(path_chk_rest) + 1
utils.load_optim(optimizer, path_chk_rest)
for i in range(1, start_epoch):
scheduler.step()
new_lr = scheduler.get_lr()[0]
print('------------------------------------------------------------------------------')
print("==> Resuming Training with learning rate:", new_lr)
print('------------------------------------------------------------------------------')
if len(device_ids)>1:
model_restoration = nn.DataParallel(model_restoration, device_ids = device_ids)
######### Loss ###########
criterion_char = losses.CharbonnierLoss()
criterion_edge = losses.EdgeLoss()
criterion_SSIM = SSIM()
crit_L1smooth = losses.L1smooth()
#train_batchsize = opt.OPTIM.BATCH_SIZE
######### DataLoaders ###########
train_dataset = get_training_data(train_dir, {'patch_size':opt.TRAINING.TRAIN_PS})
#train_loader = DataLoader(dataset=train_dataset, batch_size=opt.OPTIM.BATCH_SIZE, shuffle=True, drop_last=True, pin_memory=True)
train_loader = DataLoader(dataset=train_dataset, batch_size=opt.OPTIM.BATCH_SIZE, shuffle=True, num_workers=8, drop_last=False, pin_memory=True)
val_dataset = get_validation_data(val_dir, {'patch_size':opt.TRAINING.VAL_PS})
#val_loader = DataLoader(dataset=val_dataset, batch_size=16, shuffle=False, drop_last=True, pin_memory=True)
val_loader = DataLoader(dataset=val_dataset, batch_size=16, shuffle=False, num_workers=8, drop_last=False, pin_memory=True)
print('===> Start Epoch {} End Epoch {}'.format(start_epoch,opt.OPTIM.NUM_EPOCHS + 1))
print('===> Loading datasets')
best_psnr = 0
best_epoch = 0
for epoch in range(start_epoch, opt.OPTIM.NUM_EPOCHS + 1):
epoch_start_time = time.time()
epoch_loss = 0
SSIM_all = 0
train_id = 1
train_sample = 0
model_restoration.train()
for i, data in enumerate(tqdm(train_loader), 0):
# zero_grad
for param in model_restoration.parameters():
param.grad = None
target = data[0].cuda()
input_ = data[1].cuda()
criterion_char.cuda()
criterion_edge.cuda()
criterion_SSIM.cuda()
crit_L1smooth.cuda()
restored = model_restoration(input_)
restored[0] = restored[0].cuda()
restored[1] = restored[1].cuda()
# Compute loss at each stage
#loss_char = np.sum([criterion_char(restored[j],target) for j in range(len(restored))])
loss_char = criterion_char(restored[0],target) + 0.8*criterion_char(restored[1],target)
#loss_edge = np.sum([criterion_edge(restored[j],target) for j in range(len(restored))])
loss_edge = criterion_edge(restored[0],target) + 0.8*criterion_edge(restored[1],target)
#loss_L1smooth = np.sum([crit_L1smooth(restored[j],target) for j in range(len(restored))])
loss_L1smooth = crit_L1smooth(restored[0],target) + 0.8*crit_L1smooth(restored[1],target)
loss_SSIM = criterion_SSIM(restored[0],target)#np.sum([criterion_SSIM(restored[j],target) for j in range(len(restored))])
loss = (0.3*loss_char) + (0.2*loss_edge) - (0.15*loss_SSIM)# + (0.3*loss_L1smooth)# 0.05,0.2
loss.backward()
optimizer.step()
#scheduler.step()##############
epoch_loss +=loss.item()
SSIM_all +=loss_SSIM.item()
train_sample +=1
SSIM = SSIM_all/train_sample
if epoch ==1:
flops, params = profile(model_restoration, (input_,))
print('flops: ', flops, 'params: ', params)
a = str("[flops: %.2f --- params %d]" % (flops, params))
par_file = open(file1, 'a+')
par_file.write(a)
par_file.write('\n')
par_file.close()
#### Evaluation ####
if epoch%opt.TRAINING.VAL_AFTER_EVERY == 0:
model_restoration.eval()
psnr_val_rgb = []
for ii, data_val in enumerate((val_loader), 0):
target = data_val[0].cuda()
input_ = data_val[1].cuda()
with torch.no_grad():
restored = model_restoration(input_)
#flops, params = profile(model_restoration, (input_,))
#print('flops: ', flops, 'params: ', params)
restored = restored[0]
for res,tar in zip(restored,target):
psnr_val_rgb.append(utils.torchPSNR(res, tar))
psnr_val_rgb = torch.stack(psnr_val_rgb).mean().item()
if psnr_val_rgb > best_psnr:
best_psnr = psnr_val_rgb
best_epoch = epoch
torch.save({'epoch': epoch,
'state_dict': model_restoration.state_dict(),
'optimizer' : optimizer.state_dict()
}, os.path.join(model_dir,"model_best.pth"))
print("[epoch %d PSNR: %.4f --- best_epoch %d Best_PSNR %.4f]" % (epoch, psnr_val_rgb, best_epoch, best_psnr))
format_str1 = ('Epoch: %d, PSNR: %.4f, best_epoch: %d, Best_PSNR: %.4f')
a = str(format_str1 % (epoch, psnr_val_rgb, best_epoch, best_psnr))
PSNR_file = open(file2, 'a+')
PSNR_file.write(a)
PSNR_file.write('\n')
PSNR_file.close()
torch.save({'epoch': epoch,
'state_dict': model_restoration.state_dict(),
'optimizer' : optimizer.state_dict()
}, os.path.join(model_dir,f"model_epoch_{epoch}.pth"))
scheduler.step()
print("------------------------------------------------------------------")
print("Epoch: {}\tTime: {:.4f}\tLoss: {:.4f}\tSSIM: {:.4f}\tLearningRate {:.8f}".format(epoch, time.time()-epoch_start_time, epoch_loss, SSIM, scheduler.get_lr()[0]))
print("------------------------------------------------------------------")
#if epoch >1:
format_str = ('Epoch: %d, Time: %.4f, Loss: %.4f, SSIM: %.4f, LearningRate: %.8f')
a = str(format_str % (epoch, time.time()-epoch_start_time, epoch_loss, SSIM, scheduler.get_lr()[0]))
loss_file = open(file3, 'a+')
loss_file.write(a)
loss_file.write('\n')
loss_file.close()
torch.save({'epoch': epoch,
'state_dict': model_restoration.state_dict(),
'optimizer' : optimizer.state_dict()
}, os.path.join(model_dir,"model_latest.pth"))