forked from Denys88/rl_games
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdata.yml
82 lines (82 loc) · 1.56 KB
/
data.yml
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
algo:
name: a2c_continuous
config:
bounds_loss_coef: 0.5
clip_value: true
critic_coef: 1
e_clip: 0.2
entropy_coef: 0.0
env_name: BipedalWalkerHardcoreCnn-v3
gamma: 0.99
grad_norm: 0.5
learning_rate: 1e-4
lr_schedule: adaptive
lr_threshold: 0.008
max_epochs: 20000
mini_epochs: 4
minibatch_size: 2048
name: walker_0_1
normalize_advantage: true
normalize_input: false
num_actors: 16
ppo: true
reward_shaper:
min_val: -1
scale_value: 0.1
score_to_win: 300
seq_length: 8
steps_num: 256
tau: 0.9
truncate_grads: true
load_checkpoint: false
load_path: nn/walker_0_0ep=813.0rew=287.86985430901296
model:
name: continuous_a2c_logstd
network:
cnn:
activation: elu
convs:
- filters: 64
kernel_size: 3
padding: 1
strides: 1
- filters: 64
kernel_size: 3
padding: 1
strides: 1
- filters: 64
kernel_size: 3
padding: 1
strides: 1
- filters: 128
kernel_size: 2
padding: 0
strides: 1
initializer:
name: variance_scaling_initializer
scale: 2
regularizer:
name: None
type: conv1d
mlp:
activation: elu
initializer:
name: variance_scaling_initializer
scale: 2
regularizer:
name: None
units:
- 128
name: actor_critic
separate: true
space:
continuous:
fixed_sigma: true
mu_activation: None
mu_init:
name: variance_scaling_initializer
scale: 0.02
sigma_activation: None
sigma_init:
name: const_initializer
value: 0