Hugging Face's logo

sgoodfriend
/
dqn-CartPole-v1

Reinforcement Learning
rl-algo-impls
CartPole-v1
dqn
deep-reinforcement-learning
Eval Results
Model card Files Community
dqn-CartPole-v1
File size: 3,170 Bytes 
ff8c6a7
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
 
CartPole-v1: &cartpole-defaults
  n_timesteps: !!float 4e5
  policy_hyperparams:
    hidden_sizes: [32]
  algo_hyperparams:
    steps_per_epoch: 4096
    pi_lr: 0.01
    gamma: 0.99
    lam: 1
    val_lr: 0.01
    train_v_iters: 80
  eval_params:
    step_freq: !!float 2.5e4
    n_episodes: 10
    save_best: true

CartPole-v0:
  <<: *cartpole-defaults
  n_timesteps: !!float 1e5
  algo_hyperparams:
    steps_per_epoch: 1024
    pi_lr: 0.01
    gamma: 0.99
    lam: 1
    val_lr: 0.01
    train_v_iters: 80

Acrobot-v1:
  n_timesteps: !!float 2e5
  policy_hyperparams:
    hidden_sizes: [32, 32]
  algo_hyperparams:
    steps_per_epoch: 2048
    pi_lr: 0.005
    gamma: 0.99
    lam: 0.97
    val_lr: 0.01
    train_v_iters: 80
    max_grad_norm: 0.5
  eval_params:
    step_freq: !!float 4e4
    n_episodes: 10
    save_best: true

LunarLander-v2:
  n_timesteps: !!float 4e6
  policy_hyperparams:
    hidden_sizes: [256, 256]
  algo_hyperparams:
    steps_per_epoch: 2048
    pi_lr: 0.0001
    gamma: 0.999
    lam: 0.97
    val_lr: 0.0001
    train_v_iters: 80
    max_grad_norm: 0.5
  eval_params:
    step_freq: !!float 5e4
    n_episodes: 10
    save_best: true

CarRacing-v0:
  n_timesteps: !!float 4e6
  env_hyperparams:
    frame_stack: 4
    n_envs: 4
    vec_env_class: "dummy"
  policy_hyperparams:
    hidden_sizes: [256, 256]
  algo_hyperparams:
    steps_per_epoch: 4000
    pi_lr: !!float 7e-5
    gamma: 0.99
    lam: 0.95
    val_lr: !!float 1e-4
    train_v_iters: 40
    max_grad_norm: 0.5
  eval_params:
    step_freq: !!float 5e4
    n_episodes: 10
    save_best: true

HalfCheetahBulletEnv-v0: &pybullet-defaults
  n_timesteps: !!float 2e6
  policy_hyperparams:
    hidden_sizes: [64, 64]
    init_layers_orthogonal: true
  algo_hyperparams:
    steps_per_epoch: 4000
    pi_lr: !!float 3e-4
    gamma: 0.99
    lam: 0.97
    val_lr: !!float 1e-3
    train_v_iters: 80
    max_grad_norm: 0.5
  eval_params:
    step_freq: !!float 1e5
    n_episodes: 10
    save_best: true

HopperBulletEnv-v0:
  <<: *pybullet-defaults

AntBulletEnv-v0:
  <<: *pybullet-defaults
  policy_hyperparams:
    hidden_sizes: [400, 300]
  algo_hyperparams:
    pi_lr: !!float 7e-4
    gamma: 0.99
    lam: 0.97
    val_lr: !!float 7e-3
    train_v_iters: 80
    max_grad_norm: 0.5

FrozenLake-v1:
  n_timesteps: !!float 8e5
  env_params:
    make_kwargs:
      map_name: 8x8
      is_slippery: true
  policy_hyperparams:
    hidden_sizes: [64]
  algo_hyperparams:
    steps_per_epoch: 2048
    pi_lr: 0.01
    gamma: 0.99
    lam: 0.98
    val_lr: 0.01
    train_v_iters: 80
    max_grad_norm: 0.5
  eval_params:
    step_freq: !!float 5e4
    n_episodes: 10
    save_best: true

SpaceInvadersNoFrameskip-v4: &atari-defaults
  n_timesteps: !!float 1e7
  env_hyperparams:
    frame_stack: 4
    no_reward_timeout_steps: 1_000
    n_envs: 8
    vec_env_class: "subproc"
  policy_hyperparams:
    hidden_sizes: [256, 256]
  algo_hyperparams:
    steps_per_epoch: 4096
    pi_lr: !!float 1e-4
    gamma: 0.99
    lam: 0.95
    val_lr: !!float 2e-4
    train_v_iters: 80
    max_grad_norm: 0.5
  eval_params:
    step_freq: !!float 1e5
    n_episodes: 10
    save_best: true