-
Notifications
You must be signed in to change notification settings - Fork 24
/
Copy pathtrainer_config.yaml
118 lines (100 loc) · 2.46 KB
/
trainer_config.yaml
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
behaviors:
Fight:
trainer_type: ppo
keep_checkpoints: 10
max_steps: 1e8
time_horizon: 1000
summary_freq: 10000
threaded: true
network_settings:
normalize: false
hidden_units: 512
num_layers: 1
vis_encode_type: simple
hyperparameters:
batch_size: 1024
buffer_size: 86016 # 6 executables x 14 agents x 1024
learning_rate: 0.0003
beta: 0.01
epsilon: 0.2
lambd: 0.95
num_epoch: 3
learning_rate_schedule: linear
reward_signals:
extrinsic:
gamma: 0.99
strength: 1.0
WalkSpot:
trainer_type: ppo
keep_checkpoints: 10
max_steps: 1e8
time_horizon: 1000
summary_freq: 10000
threaded: true
network_settings:
normalize: false
hidden_units: 384 # 48 vector obs x 8
num_layers: 3
vis_encode_type: simple
hyperparameters:
batch_size: 2048
buffer_size: 73728 # 4 executables x 9 agents x 2048
learning_rate: 0.0003
beta: 0.005
epsilon: 0.2
lambd: 0.95
num_epoch: 3
learning_rate_schedule: constant
reward_signals:
extrinsic:
strength: 1.0
gamma: 0.99
# gail:
# strength: 1.0
# gamma: 0.99
# encoding_size: 128
# learning_rate: 0.0003
# use_actions: true
# use_vail: false
# demo_path: Assets/ML/Demos/Spot.demo
# behavioral_cloning:
# demo_path: Assets/ML/Demos/Spot.demo
# steps: 10000000
# strength: 0.5
WalkGunBot:
trainer_type: ppo
keep_checkpoints: 10
max_steps: 1e8
time_horizon: 1000
summary_freq: 10000
threaded: true
network_settings:
normalize: false
hidden_units: 448 # 56 vector obs x 8
num_layers: 3
vis_encode_type: simple
hyperparameters:
batch_size: 2048
buffer_size: 73728 # 4 executables x 9 agents x 2048
learning_rate: 0.0003
beta: 0.005
epsilon: 0.2
lambd: 0.95
num_epoch: 3
learning_rate_schedule: constant
reward_signals:
extrinsic:
strength: 1.0
gamma: 0.99
# gail:
# strength: 1.0
# gamma: 0.99
# encoding_size: 128
# learning_rate: 0.0003
# use_actions: true
# use_vail: false
# demo_path: Assets/ML/Demos/GunBot.demo
# behavioral_cloning:
# demo_path: Assets/ML/Demos/GunBot.demo
# steps: 10000000
# strength: 0.5