-
Notifications
You must be signed in to change notification settings - Fork 0
/
iccm2019.sh
executable file
·66 lines (55 loc) · 1.53 KB
/
iccm2019.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
#!/bin/bash
# This experiment aims to show how mean_correct_answer_step varies dependently on
# whether two operands are ordered. This means if two operands are a and b, then
# a >= b.
# Experiment name : 'ordered_operands'
# Configurations
# - operand_digits=4
# - operator='add'
# - rnn_type='jordan'
# - activation='relu'
# - hidden_units=64
# - confidence_prob=0.9
# - max_steps=30
exec &> "log_iccm2019.txt"
sleep_sec=10
experiment_name="iccm2019"
operand_digits=4
operator='add'
rnn_type='jordan'
activation='relu'
hidden_units=24
confidence_prob=0.8
max_steps=30
start_echo()
{
operator=$1
echo "==================================================================="
echo "Experiment name : ${experiment_name}"
echo "Operand digits: ${operand_digits}"
echo "Operator: ${operator}"
echo "RNN type: ${rnn_type}"
echo "Hidden activation: ${activation}"
echo "Hidden units: ${hidden_units}"
echo "Confidence probability: ${confidence_prob}"
echo "Max steps: ${max_steps}"
echo "==================================================================="
}
experiment()
{
device_num=$1
################################################################################
for i in {1..75..1}
do
start_echo $operator $rnn_type $confidence_prob
python3 rnn_run.py $experiment_name $operand_digits $operator $rnn_type $activation $hidden_units $confidence_prob $max_steps $device_num
done
}
experiment 1 &
sleep $sleep_sec
experiment 0 &
sleep $sleep_sec
experiment 1 &
sleep $sleep_sec
experiment 0 &
sleep $sleep_sec