-
Notifications
You must be signed in to change notification settings - Fork 2
/
setup.cfg
104 lines (79 loc) · 2.31 KB
/
setup.cfg
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
# main configuration to setup the network
# usage - run multiNodeLearning.py file
# to save different setting to run expect results
# The type of each node
# 0 'legacy' - Legacy (Dumb) Node
# 1 'hopping' - Hopping Node
# 2 'im' - Intermittent Node/ im Node
# 3 'dsa' - DSA node (just avoids)
# 4 'possion' - possion Node
# 5 'markovChain' - markovChain Node
# 10 'mdp' - MDP Node
# 11 'dqn' - a. DQN Node
# 12 'dqnDouble' - b. DQN-DoubleQ
# 13 'dqnPriReplay' - c. DQN-PriReplay
# 14 'dqnDuel' - d. DQN-Duel
# 15 'dqnRef' - e. DQ-Refined
# 16 'dpg' - DPG policy gradient
# 17 'ac' - Actor Critic
# 18 'ddpg' - Distributed Proximal Policy Optimization (T.B.D)
# 19 'a3cDiscrete' - A3C discrete action
# 20 'a3cDistribute'
# 21 'a3cRNN'
# 22 'dqnDynamic'
# 23 'dppo'
# 24 'et' - eligiable trace
# 25 'guess' - memory with guess sample
# partial obervation
# 30 'dqnPad' - pad till full observation DQN
# 31 'dqnPo' - shorten partial observation DQN
# 32 'dqnStack' - stacked partial obervation as input to DQN
# 33 'dpgStack' - stacked partial obervation as input to DQN
#---------------------------
# 34 'drqn' - deep recurrent q network
[Global]
numSteps = 30000
numChans = 3
optimalTP = 4
################### type of node #################################
nodeTypes = [0,0,34,34]
####################################################################
[legacyNode]
legacyChanList = [0,1]
# could be NOT 1.0, e.g. 0.1
[hoppingNode]
hoppingChanList = [ [2,3]]
hoppingWidth = 2
hopRate = 1
offSet = [50]
# the hop rate matters
[imNode]
imChanList = [2]
#imDutyCircleList = [[0.1, 0.2, 0.5, 0.9]]
imDutyCircleList = [[0.50]]
imPeriod = 5
[poissonNode]
poissonChanList = 2
arrivalInterval = 5
serviceInterval = 4
[markovChainNode]
mcChanList = 7
alpha = 0.1
beta = 0.9
[noise]
noiseErrorProb = 0.00
noiseFlipNum = 1
[partialObservation]
# for pad
poBlockNum = 4
# for partial observation
poSeeNum = 2
poStepNum = 2
padEnable = 0
padValue = 1
stackNum = 1
# To Do Pad Value
[Neural Network]
# in dqnxx.py mdp.py
# Set "explore ratio" - exploreTye, exploreDecay, exploreProbMin
# set "reward"