-
Notifications
You must be signed in to change notification settings - Fork 7
/
train.sh
116 lines (110 loc) · 3.61 KB
/
train.sh
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
source /cm/shared/engaging/anaconda/2018.12/etc/profile.d/conda.sh
conda activate neuralsym
# cannot use too high LR, will diverge slowly (loss increases > 20)
# higher bs --> faster training (using CPU)
# 8 sec/epoch on 1 RTX2080
# Highway, repeat with seed 77777777
python train.py \
--model 'Highway' \
--expt_name 'Highway_77777777_depth0_dim300_lr1e3_stop2_fac30_pat1' \
--log_file 'Highway_77777777_depth0_dim300_lr1e3_stop2_fac30_pat1' \
--do_train \
--do_test \
--prodfps_prefix 50k_1000000dim_2rad_to_32681_prod_fps \
--labels_prefix 50k_1000000dim_2rad_to_32681_labels \
--csv_prefix 50k_1000000dim_2rad_to_32681_csv \
--bs 300 \
--bs_eval 300 \
--random_seed 77777777 \
--learning_rate 1e-3 \
--epochs 30 \
--early_stop \
--early_stop_patience 2 \
--depth 0 \
--hidden_size 300 \
--lr_scheduler_factor 0.3 \
--lr_scheduler_patience 1 \
--checkpoint
# Highway, repeat with seed 20210423
# python train.py \
# --model 'Highway' \
# --expt_name 'Highway_20210423_depth0_dim300_lr1e3_stop2_fac30_pat1' \
# --log_file 'Highway_20210423_depth0_dim300_lr1e3_stop2_fac30_pat1' \
# --do_train \
# --do_test \
# --prodfps_prefix 50k_1000000dim_2rad_to_32681_prod_fps \
# --labels_prefix 50k_1000000dim_2rad_to_32681_labels \
# --csv_prefix 50k_1000000dim_2rad_to_32681_csv \
# --bs 300 \
# --bs_eval 300 \
# --random_seed 20210423 \
# --learning_rate 1e-3 \
# --epochs 30 \
# --early_stop \
# --early_stop_patience 2 \
# --depth 0 \
# --hidden_size 300 \
# --lr_scheduler_factor 0.3 \
# --lr_scheduler_patience 1 \
# --checkpoint
# FC, following Segler 2017 Neural, not better than Highway!
# python train.py \
# --model 'FC' \
# --expt_name 'FC512Elu_lr1e3_stop2_fac30_pat1' \
# --log_file 'FC512Elu_lr1e3_stop2_fac30_pat1' \
# --do_train \
# --do_test \
# --prodfps_prefix 50k_1000000dim_2rad_to_32681_prod_fps \
# --labels_prefix 50k_1000000dim_2rad_to_32681_labels \
# --csv_prefix 50k_1000000dim_2rad_to_32681_csv \
# --bs 300 \
# --bs_eval 300 \
# --random_seed 1337 \
# --learning_rate 1e-3 \
# --epochs 30 \
# --early_stop \
# --early_stop_patience 2 \
# --depth 0 \
# --hidden_size 512 \
# --lr_scheduler_factor 0.3 \
# --lr_scheduler_patience 1 \
# --checkpoint
# other experiments
# python train.py \
# --expt_name 'depth0_dim750_lr1e3_stop2_fac30_pat1' \
# --log_file 'depth0_dim750_lr1e3_stop2_fac30_pat1' \
# --do_train \
# --do_test \
# --prodfps_prefix 50k_1000000dim_2rad_to_32681_prod_fps \
# --labels_prefix 50k_1000000dim_2rad_to_32681_labels \
# --csv_prefix 50k_1000000dim_2rad_to_32681_csv \
# --bs 300 \
# --bs_eval 300 \
# --random_seed 1337 \
# --learning_rate 1e-3 \
# --epochs 30 \
# --early_stop \
# --early_stop_patience 2 \
# --depth 0 \
# --hidden_size 750 \
# --lr_scheduler_factor 0.3 \
# --lr_scheduler_patience 1
# python train.py \
# --expt_name 'depth0_dim1000_lr1e3_stop2_fac30_pat1' \
# --log_file 'depth0_dim1000_lr1e3_stop2_fac30_pat1' \
# --do_train \
# --do_test \
# --prodfps_prefix 50k_1000000dim_2rad_to_32681_prod_fps \
# --labels_prefix 50k_1000000dim_2rad_to_32681_labels \
# --csv_prefix 50k_1000000dim_2rad_to_32681_csv \
# --bs 300 \
# --bs_eval 300 \
# --random_seed 1337 \
# --learning_rate 1e-3 \
# --epochs 30 \
# --early_stop \
# --early_stop_patience 2 \
# --depth 0 \
# --hidden_size 1000 \
# --lr_scheduler_factor 0.3 \
# --lr_scheduler_patience 1