-
Notifications
You must be signed in to change notification settings - Fork 0
/
train_mae_sayavakepicutego4d.sh
executable file
·141 lines (128 loc) · 3.88 KB
/
train_mae_sayavakepicutego4d.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
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
#!/bin/bash
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=4
#SBATCH --gres=gpu:a100:4
#SBATCH --cpus-per-task=16
#SBATCH --mem=480GB
#SBATCH --time=48:00:00
#SBATCH --job-name=train_mae_sayavakepicutego4d
#SBATCH --output=train_mae_sayavakepicutego4d_%A_%a.out
#SBATCH --array=0-12
export MASTER_ADDR=$(hostname -s)
export MASTER_PORT=$(shuf -i 10000-65500 -n 1)
export WORLD_SIZE=4
DATAS=(
"sayavakepicutego4d_{000000..000017}"
"sayavakepicutego4d_0.1_1_{000000..000001}"
"sayavakepicutego4d_0.1_2_{000000..000001}"
"sayavakepicutego4d_0.1_3_{000000..000001}"
"sayavakepicutego4d_0.01_1_000000"
"sayavakepicutego4d_0.01_2_000000"
"sayavakepicutego4d_0.01_3_000000"
"sayavakepicutego4d_0.001_1_000000"
"sayavakepicutego4d_0.001_2_000000"
"sayavakepicutego4d_0.001_3_000000"
"sayavakepicutego4d_0.0001_1_000000"
"sayavakepicutego4d_0.0001_2_000000"
"sayavakepicutego4d_0.0001_3_000000"
)
SAVES=(
"sayavakepicutego4d"
"sayavakepicutego4d_0.1_1"
"sayavakepicutego4d_0.1_2"
"sayavakepicutego4d_0.1_3"
"sayavakepicutego4d_0.01_1"
"sayavakepicutego4d_0.01_2"
"sayavakepicutego4d_0.01_3"
"sayavakepicutego4d_0.001_1"
"sayavakepicutego4d_0.001_2"
"sayavakepicutego4d_0.001_3"
"sayavakepicutego4d_0.0001_1"
"sayavakepicutego4d_0.0001_2"
"sayavakepicutego4d_0.0001_3"
)
DATA=${DATAS[$SLURM_ARRAY_TASK_ID]}
SAVE=${SAVES[$SLURM_ARRAY_TASK_ID]}
echo $DATA
echo $SAVE
# # vit-s/14
# srun python -u /scratch/eo41/mae/train_mae.py \
# --model 'mae_vit_small_patch14' \
# --resume "" \
# --batch_size_per_gpu 512 \
# --mask_ratio 0.8 \
# --num_workers 16 \
# --lr 0.0003 \
# --min_lr 0.0003 \
# --weight_decay 0.0 \
# --output_dir "/vast/eo41/sayavakepicutego4d_models/mae_vits14" \
# --data_path "/vast/eo41/sayavakepicutego4d/${DATA}.tar" \
# --save_prefix "${SAVE}_vits14"
# # vit-b/14
# srun python -u /scratch/eo41/mae/train_mae.py \
# --model 'mae_vit_base_patch14' \
# --resume "" \
# --batch_size_per_gpu 256 \
# --mask_ratio 0.8 \
# --num_workers 16 \
# --lr 0.0003 \
# --min_lr 0.0003 \
# --weight_decay 0.0 \
# --output_dir "/vast/eo41/sayavakepicutego4d_models/mae_vitb14" \
# --data_path "/vast/eo41/sayavakepicutego4d/${DATA}.tar" \
# --save_prefix "${SAVE}_vitb14"
# # vit-l/14
# srun python -u /scratch/eo41/mae/train_mae.py \
# --model 'mae_vit_large_patch14' \
# --resume "" \
# --batch_size_per_gpu 256 \
# --mask_ratio 0.8 \
# --num_workers 16 \
# --lr 0.0003 \
# --min_lr 0.0003 \
# --weight_decay 0.0 \
# --output_dir "/vast/eo41/sayavakepicutego4d_models/mae_vitl14" \
# --data_path "/vast/eo41/sayavakepicutego4d/${DATA}.tar" \
# --save_prefix "${SAVE}_vitl14"
# # vit-h/14
# srun python -u /scratch/eo41/mae/train_mae.py \
# --model 'mae_vit_huge_patch14' \
# --resume "" \
# --mask_ratio 0.8 \
# --batch_size_per_gpu 256 \
# --num_workers 16 \
# --lr 0.0001 \
# --min_lr 0.0001 \
# --weight_decay 0.0 \
# --output_dir "/vast/eo41/sayavakepicutego4d_models/mae_vith14" \
# --data_path "/vast/eo41/sayavakepicutego4d/${DATA}.tar" \
# --save_prefix "${SAVE}_vith14"
# # vit-h/14 @ 448px
# srun python -u /scratch/eo41/mae/train_mae.py \
# --model 'mae_vit_huge_patch14_448' \
# --resume "" \
# --input_size 448 \
# --mask_ratio 0.8 \
# --batch_size_per_gpu 42 \
# --num_workers 16 \
# --lr 0.0001 \
# --min_lr 0.0001 \
# --weight_decay 0.0 \
# --output_dir "/vast/eo41/sayavakepicutego4d_models/mae_vith14_448" \
# --data_path "/vast/eo41/sayavakepicutego4d/${DATA}.tar" \
# --save_prefix "${SAVE}_vith14_448"
# vit-h/14 @ 476 px
srun python -u /scratch/eo41/mae/train_mae.py \
--model 'mae_vit_huge_patch14_476' \
--resume "" \
--input_size 476 \
--mask_ratio 0.8 \
--batch_size_per_gpu 34 \
--num_workers 16 \
--lr 0.0001 \
--min_lr 0.0001 \
--weight_decay 0.0 \
--output_dir "/vast/eo41/sayavakepicutego4d_models/mae_vith14_476" \
--data_path "/vast/eo41/sayavakepicutego4d/${DATA}.tar" \
--save_prefix "${SAVE}_vith14_476"
echo "Done"