-
Notifications
You must be signed in to change notification settings - Fork 8
/
incremental_procedure_buffer.py
191 lines (153 loc) · 7.31 KB
/
incremental_procedure_buffer.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
# coding=utf-8
from __future__ import absolute_import, print_function
import argparse
import datetime
import logging
import math
import random
import os
import time
import torch
from utils import get_time_str
from os import path as osp
import numpy as np
from copy import deepcopy
from data import create_dataloader, create_dataset, create_sampler
from methods import create_model
from utils.options import dict2str, parse
from utils import (MessageLogger, get_env_info, get_root_logger,
init_tb_logger, init_wandb_logger, check_resume,
make_exp_dirs, set_random_seed, set_gpu, Averager)
def main():
parser = argparse.ArgumentParser()
parser.add_argument(
'-opt',type=str, required=True, help='Path to option YAML file.')
parser.add_argument('--local_rank', type=int, default=0)
args = parser.parse_args()
opt = parse(args.opt, is_train=False, is_incremental=True)
rank = 0
world_size = 1
opt['rank'] = 0
opt['world_size'] = 1
make_exp_dirs(opt)
log_file = osp.join(opt['path']['log'],
f"incremental_{opt['name']}_{get_time_str()}.log")
logger = get_root_logger(
logger_name='FS-IL', log_level=logging.INFO, log_file=log_file)
logger.info(get_env_info())
logger.info(dict2str(opt))
# initialize tensorboard logger and wandb logger
tb_logger = None
if opt['logger'].get('use_tb_logger') and 'debug' not in opt['name']:
log_dir = './tb_logger/' + opt['name']
tb_logger = init_tb_logger(log_dir=log_dir)
if (opt['logger'].get('wandb')
is not None) and (opt['logger']['wandb'].get('project')
is not None) and ('debug' not in opt['name']):
assert opt['logger'].get('use_tb_logger') is True, (
'should turn on tensorboard when using wandb')
wandb_logger = init_wandb_logger(opt)
else:
wandb_logger = None
opt['wandb_logger'] = wandb_logger
# set gpu
# set_gpu(opt['gpu'])
# set random seed
seed = opt['manual_seed']
if seed is None:
seed = random.randint(1, 10000)
opt['manual_seed'] = seed
logger.info(f'Random seed: {seed}')
set_random_seed(seed + rank)
torch.backends.cudnn.benchmark = True
# define the variables for incremental few-shot learning
total_classes = opt['datasets']['train']['total_classes']
bases = opt['train']['bases']
num_tasks = opt['train']['tasks']
num_shots = opt['train']['shots']
fine_tune = opt['train']['fine_tune']
fine_tune_epoch = opt['train']['fine_tune_epoch']
num_class_per_task = int((total_classes - bases) / (num_tasks - 1))
opt['train']['num_class_per_task'] = num_class_per_task
if opt.get('Random', True):
random_class_perm = np.random.permutation(total_classes)
else:
random_class_perm = np.arange(total_classes)
opt['class_permutation'] = random_class_perm
# deep copy the opt
opt_old = deepcopy(opt)
num_tests = opt['train']['num_test']
acc_avg = [Averager() for i in range(num_tasks)]
for test_id in range(num_tests):
opt = deepcopy(opt_old)
for task_id in range(num_tasks):
opt['test_id'] = test_id
# deep copy the opt
# Load the model of former session
# 'task_id = -1' indicates that the program will not load the prototypes, and just load the base model
opt['task_id'] = task_id - 1
model = create_model(opt)
opt['task_id'] = task_id
val_classes = random_class_perm[:bases + task_id * num_class_per_task]
if task_id == 0:
selected_classes = random_class_perm[:bases]
else:
selected_classes = random_class_perm[bases + (task_id - 1) * num_class_per_task:
bases + task_id * num_class_per_task]
IL = opt['train'].get('IL', False)
# creating the dataset
for phase, dataset_opt in opt['datasets'].items():
if phase == 'train':
dataset_opt['all_classes'] = random_class_perm
dataset_opt['selected_classes'] = selected_classes
train_set = create_dataset(dataset_opt=dataset_opt)
if task_id > 0 and not IL:
session_path_root, _ = os.path.split(dataset_opt['dataroot'])
Random = opt.get('Random', True)
if opt['manual_seed'] != 1997:
seed = opt['manual_seed']
index_root = osp.join(session_path_root,
f'Random{Random}_seed{seed}_bases{bases}_tasks{num_tasks}_shots{num_shots}',
f'test_{test_id}', f'session_{task_id}', 'index.pt')
else:
index_root = osp.join(session_path_root,
f'Random{Random}_seed{seed}_bases{bases}_tasks{num_tasks}_shots{num_shots}',
f'test_{test_id}', f'session_{task_id}', 'index.pt')
index = torch.load(index_root)
train_set.sample_the_buffer_data_with_index(index)
if phase == 'val':
dataset_opt['all_classes'] = random_class_perm
dataset_opt['selected_classes'] = val_classes
val_set = create_dataset(dataset_opt=dataset_opt)
model.incremental_init(train_set, val_set)
if task_id > 0 and fine_tune:
tb_logger_temp = tb_logger if test_id == 0 else None
model.incremental_fine_tune(train_dataset=train_set, val_dataset=val_set,
num_epoch=fine_tune_epoch, task_id=task_id, test_id=test_id,
tb_logger=tb_logger_temp)
print('fine-tune procedure is finished!')
# update incremental setting
model.incremental_update(train_set)
acc = model.incremental_test(val_set, task_id, test_id)
acc_avg[task_id].add(acc)
model.save(epoch=-1, current_iter=task_id, name=f'test{test_id}_session', dataset=train_set)
opt = deepcopy(opt_old)
model.set_the_saving_files_path(opt=opt, task_id=task_id)
print(f'Successfully saving the model of test {test_id} session {task_id}')
message = f'--------------------------Final Avg Acc-------------------------'
logger.info(message)
for i, acc in enumerate(acc_avg):
data = acc.obtain_data()
m = np.mean(data)
std = np.std(data)
pm = 1.96 * (std / np.sqrt(len(data)))
message = f'Session {i + 1}: {m*100:.2f}+-{pm*100:.2f}'
logger.info(message)
if tb_logger:
tb_logger.add_scalar(f'sessions_acc', acc.item(), i)
if wandb_logger is not None:
wandb_logger.log({f'sessions_acc': acc.item()}, step=i)
print('finish!!')
print('finish')
if __name__ == '__main__':
main()