forked from chainer/chainerrl
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtest_examples.sh
executable file
·48 lines (37 loc) · 1.99 KB
/
test_examples.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
#!/bin/bash
set -Ceu
outdir=$(mktemp -d)
echo "outdir: $outdir"
gpu="$1"
# ale/dqn
python examples/ale/train_dqn_ale.py pong --steps 100 --replay-start-size 50 --outdir $outdir/ale/dqn --gpu $gpu
model=$(find $outdir/ale/dqn -name "*_finish")
python examples/ale/train_dqn_ale.py pong --demo --load $model --eval-n-runs 1 --gpu $gpu
# ale/a3c
python examples/ale/train_a3c_ale.py 4 pong --steps 100 --outdir $outdir/ale/a3c
model=$(find $outdir/ale/a3c -name "*_finish")
python examples/ale/train_a3c_ale.py 4 pong --demo --load $model --eval-n-runs 1
# ale/acer
python examples/ale/train_acer_ale.py 4 pong --steps 100 --outdir $outdir/ale/acer
model=$(find $outdir/ale/acer -name "*_finish")
python examples/ale/train_acer_ale.py 4 pong --demo --load $model --eval-n-runs 1
# ale/nsq
python examples/ale/train_nsq_ale.py 4 pong --steps 100 --outdir $outdir/ale/nsq
model=$(find $outdir/ale/nsq -name "*_finish")
python examples/ale/train_nsq_ale.py 4 pong --demo --load $model --eval-n-runs 1
# gym/dqn
python examples/gym/train_dqn_gym.py --steps 100 --replay-start-size 50 --outdir $outdir/gym/dqn --gpu $gpu
model=$(find $outdir/gym/dqn -name "*_finish")
python examples/gym/train_dqn_gym.py --demo --load $model --eval-n-runs 1 --gpu $gpu
# gym/a3c
python examples/gym/train_a3c_gym.py 4 --steps 100 --outdir $outdir/gym/a3c
model=$(find $outdir/gym/a3c -name "*_finish")
python examples/gym/train_a3c_gym.py 4 --demo --load $model --eval-n-runs 1
# gym/acer
python examples/gym/train_acer_gym.py 4 --steps 100 --outdir $outdir/gym/acer
model=$(find $outdir/gym/acer -name "*_finish")
python examples/gym/train_acer_gym.py 4 --demo --load $model --eval-n-runs 1
# gym/ddpg (specify non-mujoco env to test without mujoco)
python examples/gym/train_ddpg_gym.py --steps 100 --replay-start-size 50 --minibatch-size 32 --outdir $outdir/gym/ddpg --env Pendulum-v0 --gpu $gpu
model=$(find $outdir/gym/ddpg -name "*_finish")
python examples/gym/train_ddpg_gym.py --demo --load $model --eval-n-runs 1 --env Pendulum-v0 --gpu $gpu