forked from PatrickHua/SimSiam
/
main.py
161 lines (116 loc) · 5.59 KB
/
main.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
import os
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchvision
import numpy as np
from tqdm import tqdm
from arguments import get_args
from augmentations import get_aug
from models import get_model
from tools import AverageMeter, knn_monitor, Logger, file_exist_check
from datasets import get_dataset
from optimizers import get_optimizer, LR_Scheduler
from linear_eval import main as linear_eval
from datetime import datetime
# distributed training
import torch.distributed as dist
import torch.multiprocessing as mp
from torch.nn.parallel import DistributedDataParallel as DDP
# CUDA_VISIBLE_DEVICES=2,3 python main.py --data_dir ./data/ --log_dir ./logs/ -c configs/simsiam_cifar.yaml --ckpt_dir ./cache/ --hide_progress
def main(gpu, args):
rank = args.nr * args.gpus + gpu
dist.init_process_group("nccl", rank=rank, world_size=args.world_size)
torch.manual_seed(0)
torch.cuda.set_device(gpu)
train_dataset = get_dataset(transform=get_aug(train=True, **args.aug_kwargs), train=True, **args.dataset_kwargs)
train_sampler = torch.utils.data.distributed.DistributedSampler(train_dataset, num_replicas=args.world_size, rank=rank)
train_loader = torch.utils.data.DataLoader(
dataset=train_dataset,
shuffle=False,
batch_size=(args.train.batch_size // args.gpus),
sampler = train_sampler,
**args.dataloader_kwargs
)
memory_dataset = get_dataset(transform=get_aug(train=False,train_classifier=False, **args.aug_kwargs), train=True, **args.dataset_kwargs)
memory_loader = torch.utils.data.DataLoader(
dataset=memory_dataset,
shuffle=False,
batch_size=(args.train.batch_size // args.gpus),
**args.dataloader_kwargs
)
test_datset = get_dataset( transform=get_aug(train=False, train_classifier=False, **args.aug_kwargs), train=False,**args.dataset_kwargs)
test_loader = torch.utils.data.DataLoader(
dataset= test_datset,
shuffle=False,
batch_size=(args.train.batch_size // args.gpus),
**args.dataloader_kwargs
)
print("Batch size:",(args.train.batch_size // args.gpus))
# define model
model = get_model(args.model).cuda(gpu)
model = torch.nn.SyncBatchNorm.convert_sync_batchnorm(model)
model = DDP(model, device_ids=[gpu], find_unused_parameters=True)
# define optimizer
optimizer = get_optimizer(
args.train.optimizer.name, model,
lr=args.train.base_lr*args.train.batch_size/256,
momentum=args.train.optimizer.momentum,
weight_decay=args.train.optimizer.weight_decay)
lr_scheduler = LR_Scheduler(
optimizer,
args.train.warmup_epochs, args.train.warmup_lr*args.train.batch_size/256,
args.train.num_epochs, args.train.base_lr*args.train.batch_size/256, args.train.final_lr*args.train.batch_size/256,
len(train_loader),
constant_predictor_lr=True # see the end of section 4.2 predictor
)
if gpu ==0:
logger = Logger(tensorboard=args.logger.tensorboard, matplotlib=args.logger.matplotlib, log_dir=args.log_dir)
accuracy = 0
# Start training
global_progress = tqdm(range(0, args.train.stop_at_epoch), desc=f'Training')
for epoch in global_progress:
model.train()
local_progress=tqdm(train_loader, desc=f'Epoch {epoch}/{args.train.num_epochs}', disable=args.hide_progress)
for idx, ((images1, images2), labels) in enumerate(local_progress):
model.zero_grad()
data_dict = model.forward(images1.cuda(non_blocking=True), images2.cuda(non_blocking=True))
loss = data_dict['loss'] # ddp
loss.backward()
optimizer.step()
lr_scheduler.step()
data_dict.update({'lr':lr_scheduler.get_lr()})
local_progress.set_postfix(data_dict)
if gpu ==0:
logger.update_scalers(data_dict)
if args.train.knn_monitor and epoch % args.train.knn_interval == 0 and gpu==0:
accuracy = knn_monitor(model.module.backbone, memory_loader, test_loader, gpu, k=min(args.train.knn_k, len(memory_loader.dataset)), hide_progress=args.hide_progress)
epoch_dict = {"epoch":epoch, "accuracy":accuracy}
global_progress.set_postfix(epoch_dict)
if gpu == 0:
logger.update_scalers(epoch_dict)
# Save checkpoint
if gpu ==0 and epoch % args.train.knn_interval == 0:
model_path = os.path.join(args.ckpt_dir, f"{args.name}_{epoch+1}.pth") # datetime.now().strftime('%Y%m%d_%H%M%S')
torch.save({
'epoch': epoch+1,
'state_dict':model.module.state_dict()
}, model_path)
print(f"Model saved to {model_path}")
with open(os.path.join(args.log_dir, f"checkpoint_path.txt"), 'w+') as f:
f.write(f'{model_path}')
# if args.eval is not False and gpu == 0:
# args.eval_from = model_path
# linear_eval(args)
dist.destroy_process_group()
if __name__ == "__main__":
args = get_args()
os.environ["MASTER_ADDR"] = "127.0.0.1"
os.environ["MASTER_PORT"] = "8000"
args.world_size = args.gpus * args.nodes
# Initialize the process and join up with the other processes.
# This is “blocking,” meaning that no process will continue until all processes have joined.
mp.spawn(main, args=(args,), nprocs=args.gpus, join=True)
completed_log_dir = args.log_dir.replace('in-progress', 'debug' if args.debug else 'completed')
os.rename(args.log_dir, completed_log_dir)
print(f'Log file has been saved to {completed_log_dir}')