-
Notifications
You must be signed in to change notification settings - Fork 0
/
config.yaml
91 lines (71 loc) · 2.06 KB
/
config.yaml
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
########### 說明區 ##########
#
#
#
#############################
hp:
# types of task: S/P/R/S_Mixup/R_Mixup/R_SA_Mixup
task: 'R_SA_Mixup'
# Dataset: wm811k/mitbih
dataset: 'mitbih'
repeat_times: 5
#
# s model
n_epoch_for_s_model: -1 #300
max_esc_for_s_model: -1 #100
# pl model
n_epoch_for_pl_model: 300
max_esc_for_pl_model: 300
# Pseudolabeling
pl_rate: 0.5
pl_round: 5
pl_max_esc: 5
pl_n_limit: 10000
pl_threshold: 0
# optimizer:adam/Ranger/SGD, lr:0.0001/0.001/0.1
optimizer: 'Ranger'
lr: 0.001
weight_decay: 0.0005
drop_out: 0.2
momentum: 0.9
# Batch
accumulate_gradient: True
train_batch: 4
val_batch: 4
# matters if only accumulate_gradient is true, should be a dividend of train/val batch.
train_batch_after_accumulate: 64
n_worker: 2
w_wm811k: 32
h_wm811k: 32
w_mitbih: 128
h_mitbih: 128
log: true
num_pl_epoch: 10
num_epoch: 300
num_round: 5
esc_round: 5
max_esc: 300
alpha: 0.2
beta: 0.8
pretrained: False
path:
# Wide-Resnet
# s_model_pretrained: 'models_pretrained/s_model_balance.pth'
# # WM811K
# img_list_wm811k: 'data/wm811k/train_3150_data.npy'
# indice_list_wm811k: 'data/wm811k/train_3150_10p_label_indices.npy'
# label_list_wm811k: 'data/wm811k/train_3150_10p_label_values.npy'
# test_img_list_wm811k: 'data/wm811k/test_700_data.npy'
# test_indice_list_wm811k: 'data/wm811k/test_700_label_indices.npy'
# test_label_list_wm811k: 'data/wm811k/test_700_label_values.npy'
# MITBIH
# img_list_mitbih: 'data/MITBIH/train_data_2D_2000.npy'
# img_list_mitbih: 'data/MITBIH2/train_data_2D_2000.npy'
# # Testing data
# # test_img_list_mitbih: 'data/MITBIH/test_data_2D_500.npy'
# test_img_list_mitbih: 'data/MITBIH2/test_data_2D_500.npy'
# # test_label_list_mitbih: 'data/MITBIH/test_2D_500_label_values.npy'
# test_label_list_mitbih: 'data/MITBIH2/test_2D_500_label_values.npy'
# traintools.py, loaders.py -> revise path to this file if needed
BestVal_each_round:
TestAcc_each_round: