-
Notifications
You must be signed in to change notification settings - Fork 3
/
exec_dagnn.sh
45 lines (36 loc) · 2.42 KB
/
exec_dagnn.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
#!/bin/bash
n_times=100
n_agents=50
n_exp=15
start_idx=0
filterLength=3
n_vis=24
radius=1.5
vinit=1.0
root_path='/ssd1/tkhu/AirSim/optimal_vinit_'$vinit'_R_'$radius
mode='optimal'
arch='vis_dagnn'
comm_model='knn'
K_neighbor=10
#python grad_train_data_optimal_location.py --n-times $n_times --n-exp $n_exp --start-idx $start_idx --K $filterLength --exp-name $root_path --mode $mode --radius $radius --K-neighbor $K_neighbor
python prepare_training.py --n-times $n_times --n-agents $n_agents --n-exp $n_exp --start-idx $start_idx --K $filterLength --exp-name $root_path --K $filterLength --radius $radius --vinit $vinit --arch $arch --F $n_vis --comm-model $comm_model --mode $mode --K-neighbor $K_neighbor
CUDA_VISIBLE_DEVICES=0 python train.py $arch --K $filterLength --radius $radius --vinit $vinit --F $n_vis --comm-model $comm_model --mode $mode --K-neighbor $K_neighbor
if [[ $comm_model == "disk" ]]
then
model_path='checkpoint_all_'$mode'_'$arch'_K_'$filterLength'_n_vis_'$n_vis'_R_'$radius'_vinit_'$vinit'_comm_model_'$comm_model'.tar'
root_path_aug='/ssd1/tkhu/AirSim/'$arch'_K_'$filterLength'_n_vis_'$n_vis'_R_'$radius'_vinit_'$vinit'_comm_model_'$comm_model
else
model_path='checkpoint_all_'$mode'_'$arch'_K_'$filterLength'_n_vis_'$n_vis'_vinit_'$vinit'_comm_model_'$comm_model'_K_neighbor_'$K_neighbor'.tar'
root_path_aug='/ssd1/tkhu/AirSim/'$arch'_K_'$filterLength'_n_vis_'$n_vis'_vinit_'$vinit'_comm_model_'$comm_model'_K_neighbor_'$K_neighbor
fi
mode='vis_dagnn'
n_exp_aug=5
start_idx_aug=0
echo $root_path
echo $model_path
CUDA_VISIBLE_DEVICES=0 python grad_train_data_optimal_location.py --n-times $n_times --n-exp $n_exp_aug --start-idx $start_idx_aug --K $filterLength --radius $radius --vinit $vinit --arch $arch --F $n_vis --exp-name $root_path_aug --mode $mode --model-path $model_path --comm-model $comm_model --K-neighbor $K_neighbor
joint_exp_name=$root_path','$root_path_aug
joint_start_idx=$start_idx','$start_idx_aug
joint_n_exp=$n_exp','$n_exp_aug
python prepare_training.py --n-times $n_times --n-agents $n_agents --n-exp $joint_n_exp --start-idx $joint_start_idx --K $filterLength --exp-name $joint_exp_name --K $filterLength --radius $radius --vinit $vinit --arch $arch --F $n_vis --comm-model $comm_model --mode $mode --K-neighbor $K_neighbor
CUDA_VISIBLE_DEVICES=0 python train.py $arch --K $filterLength --radius $radius --vinit $vinit --F $n_vis --comm-model $comm_model --mode $mode --K-neighbor $K_neighbor