-
Notifications
You must be signed in to change notification settings - Fork 4
/
main.py
258 lines (215 loc) · 11.7 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
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
# ------------------------------------------------------------------------
# INTR
# Copyright (c) 2023 Imageomics Paul. All Rights Reserved.
# Licensed under the Apache License, Version 2.0 [see LICENSE for details]
# ------------------------------------------------------------------------
# Copied from DETR (https://github.com/facebookresearch/detr)
# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
# ------------------------------------------------------------------------
import os
import json
import time
import random
import datetime
import argparse
import torch
import torch.nn as nn
from torch.utils.data import DataLoader, DistributedSampler
import numpy as np
from pathlib import Path
import datasets
import util.misc as utils
from models import build_model
from datasets import build_dataset
from engine import evaluate, train_one_epoch
def get_args_parser():
parser = argparse.ArgumentParser('Set transformer detector', add_help=False)
parser.add_argument('--lr', default=5.00e-5, type=float)
parser.add_argument('--lr_backbone', default=1.00e-5, type=float)
parser.add_argument('--min_lr', default=1.00e-6, type=float)
parser.add_argument('--weight_decay', default=1e-6, type=float)
parser.add_argument('--batch_size', default=12, type=int)
parser.add_argument('--epochs', default=140, type=int)
parser.add_argument('--lr_drop', default=80, type=int)
parser.add_argument('--lr_scheduler', default="StepLR", type=str, choices=["StepLR", "CosineAnnealingLR"])
parser.add_argument('--clip_max_norm', default=0.1, type=float,
help='gradient clipping max norm')
# * Backbone parameters
parser.add_argument('--backbone', default='resnet50', type=str,
help="Name of the convolutional backbone to use")
parser.add_argument('--dilation', action='store_true',
help="If true, we replace stride with dilation in the last convolutional block (DC5)")
parser.add_argument('--position_embedding', default='sine', type=str, choices=('sine', 'learned'),
help="Type of positional embedding to use on top of the image features")
# * Transformer parameters
parser.add_argument('--enc_layers', default=6, type=int,
help="Number of encoding layers in the transformer")
parser.add_argument('--dec_layers', default=6, type=int,
help="Number of decoding layers in the transformer")
parser.add_argument('--dim_feedforward', default=2048, type=int,
help="Intermediate size of the feedforward layers in the transformer blocks")
parser.add_argument('--hidden_dim', default=256, type=int,
help="Size of the embeddings (dimension of the transformer)")
parser.add_argument('--dropout', default=0.1, type=float,
help="Dropout applied in the transformer")
parser.add_argument('--nheads', default=8, type=int,
help="Number of attention heads inside the transformer's attentions")
parser.add_argument('--num_queries', default=200, type=int,
help="Number of query slots which equals to number of classes")
parser.add_argument('--pre_norm', action='store_true')
# * Dataset parameters
parser.add_argument('--dataset_name', default='cub', type=str)
parser.add_argument('--dataset_path', default='/path/to/datasets', type=str)
parser.add_argument('--output_dir', default='output',
help='path where to save, empty for no saving')
parser.add_argument('--output_sub_dir', default='output_sub')
# * INTR parameters
parser.add_argument('--noise_frac', default=0.1, type=float,
help='fraction of noise to be added to new queries while loading pretrained model')
# parser.add_argument('--rm_freeze', default=140, type=int, help='epoch at which the freezing at the encoder is removed')
parser.add_argument('--test', default="val", type=str, choices=["val", "test"])
parser.add_argument('--resume', default='', help='resume from checkpoint')
parser.add_argument('--finetune', default='',
help='finetune from pretrained checkpoint (COCO dataset trained for object detection task)')
# * Device parameters
parser.add_argument('--device', default='cuda',
help='device to use for training / testing')
parser.add_argument('--seed', default=42, type=int)
parser.add_argument('--start_epoch', default=0, type=int, metavar='N',
help='start epoch')
parser.add_argument('--eval', action='store_true')
# * Distributed training parameters
parser.add_argument('--num_workers', default=2, type=int)
parser.add_argument('--world_size', default=1, type=int,
help='number of distributed processes')
parser.add_argument('--dist_url', default='env://', help='url used to set up distributed training')
return parser
def main(args):
utils.init_distributed_mode(args)
print("git:\n {}\n".format(utils.get_sha()))
print(args)
device = torch.device(args.device)
# fix the seed for reproducibility
seed = args.seed + utils.get_rank()
torch.manual_seed(seed)
np.random.seed(seed)
random.seed(seed)
model, criterion= build_model(args)
model.to(device)
model_without_ddp = model
if args.distributed:
model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.gpu])
## for 2-phase training
# model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.gpu], find_unused_parameters=True)
model_without_ddp = model.module
n_parameters = sum(p.numel() for p in model.parameters() if p.requires_grad)
print('number of params:', n_parameters)
param_dicts = [
{"params": [p for n, p in model_without_ddp.named_parameters() if "backbone" not in n and p.requires_grad]},
{
"params": [p for n, p in model_without_ddp.named_parameters() if "backbone" in n and p.requires_grad],
"lr": args.lr_backbone,
},
]
if args.lr_scheduler=="StepLR":
optimizer = torch.optim.AdamW(param_dicts, lr=args.lr,
weight_decay=args.weight_decay)
lr_scheduler = torch.optim.lr_scheduler.StepLR(optimizer, args.lr_drop)
if args.lr_scheduler=="CosineAnnealingLR":
optimizer = torch.optim.AdamW(param_dicts, lr=args.lr,
weight_decay=args.weight_decay)
lr_scheduler = torch.optim.lr_scheduler.CosineAnnealingLR(optimizer, T_max=args.epochs, eta_min=args.min_lr)
dataset_train = build_dataset(image_set='train', args=args)
dataset_val = build_dataset(image_set=args.test, args=args)
if args.distributed:
sampler_train = DistributedSampler(dataset_train)
sampler_val = DistributedSampler(dataset_val, shuffle=False)
else:
sampler_train = torch.utils.data.RandomSampler(dataset_train)
sampler_val = torch.utils.data.SequentialSampler(dataset_val)
batch_sampler_train = torch.utils.data.BatchSampler(
sampler_train, args.batch_size, drop_last=True)
data_loader_train = DataLoader(dataset_train, batch_sampler=batch_sampler_train,
collate_fn=utils.collate_fn, num_workers=args.num_workers)
data_loader_val = DataLoader(dataset_val, args.batch_size, sampler=sampler_val,
drop_last=False, collate_fn=utils.collate_fn, num_workers=args.num_workers)
# We create output directories to store results
output_dir = Path(args.output_dir)
if not os.path.exists(os.path.join(output_dir, args.dataset_name)):
os.makedirs(os.path.join(output_dir, args.dataset_name), exist_ok=True)
if not os.path.exists(os.path.join(output_dir, args.dataset_name, args.output_sub_dir)):
os.makedirs(os.path.join(output_dir, args.dataset_name, args.output_sub_dir), exist_ok=True)
if args.resume:
if args.resume.startswith('https'):
checkpoint = torch.hub.load_state_dict_from_url(
args.resume, map_location='cpu', check_hash=True)
else:
checkpoint = torch.load(args.resume, map_location='cpu')
model_without_ddp.load_state_dict(checkpoint['model'])
if not args.eval and 'optimizer' in checkpoint and 'lr_scheduler' in checkpoint and 'epoch' in checkpoint:
optimizer.load_state_dict(checkpoint['optimizer'])
lr_scheduler.load_state_dict(checkpoint['lr_scheduler'])
args.start_epoch = checkpoint['epoch'] + 1
if args.eval:
test_stats = evaluate(model, criterion,
data_loader_val, device, args.output_dir)
if args.output_dir and utils.is_main_process():
with (output_dir / args.dataset_name / args.output_sub_dir/ "log.txt").open("a") as f:
f.write(json.dumps(test_stats) + "\n")
return
if args.finetune:
if args.finetune.startswith('https'):
checkpoint = torch.hub.load_state_dict_from_url(
args.finetune, map_location='cpu', check_hash=True)
else:
checkpoint = torch.load(args.finetune, map_location='cpu')
state_dict = checkpoint['model']
state_dict=utils.load_model(args, state_dict)
model_without_ddp.load_state_dict(state_dict)
for param in model_without_ddp.parameters():
param.requires_grad = True
model_without_ddp.to(device)
print("Start training")
start_time = time.time()
for epoch in range(args.start_epoch, args.epochs):
## for 2-phase training
# if epoch>=args.rm_freeze:
# for param in model_without_ddp.transformer.encoder.parameters():
# param.requires_grad = True
if args.distributed:
sampler_train.set_epoch(epoch)
train_stats = train_one_epoch(
model, criterion, data_loader_train, optimizer, device, epoch,
args.clip_max_norm)
lr_scheduler.step()
if args.output_dir:
checkpoint_paths = [output_dir / args.dataset_name / args.output_sub_dir/ 'checkpoint.pth']
if (epoch + 1) % args.lr_drop == 0 or (epoch + 1)==args.epochs:
checkpoint_paths.append(output_dir / args.dataset_name / args.output_sub_dir / f'checkpoint{epoch:04}.pth')
for checkpoint_path in checkpoint_paths:
utils.save_on_master({
'model': model_without_ddp.state_dict(),
'optimizer': optimizer.state_dict(),
'lr_scheduler': lr_scheduler.state_dict(),
'epoch': epoch,
'args': args,
}, checkpoint_path)
test_stats = evaluate(
model, criterion, data_loader_val, device, args.output_dir
)
log_stats = {**{f'train_{k}': v for k, v in train_stats.items()},
**{f'test_{k}': v for k, v in test_stats.items()},
'epoch': epoch,
'n_parameters': n_parameters}
if args.output_dir and utils.is_main_process():
with (output_dir / args.dataset_name / args.output_sub_dir/ "log.txt").open("a") as f:
f.write(json.dumps(log_stats) + "\n")
total_time = time.time() - start_time
total_time_str = str(datetime.timedelta(seconds=int(total_time)))
print('Training time {}'.format(total_time_str))
if __name__ == '__main__':
parser = argparse.ArgumentParser('INTR training and evaluation script', parents=[get_args_parser()])
args = parser.parse_args()
if args.output_dir:
Path(args.output_dir).mkdir(parents=True, exist_ok=True)
main(args)