ML-Dungeon/configuration.yaml at main · StevenRice99/ML-Dungeon · GitHub

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
behaviors:
  DungeonPlayer:
    trainer_type: ppo
    hyperparameters:
      batch_size: 1024
      buffer_size: 20480
      learning_rate: 3.0e-4
      learning_rate_schedule: linear
      beta: 1.0e-2
      beta_schedule: linear
      epsilon: 0.2
      epsilon_schedule: linear
      lambd: 0.95
      num_epoch: 3
      shared_critic: false
    network_settings:
      vis_encode_type: match3
      normalize: true
      hidden_units: 256
      num_layers: 2
    reward_signals:
      extrinsic:
        gamma: 0.99
        strength: 1.0
      curiosity:
        gamma: 0.99
        strength: 0.1
        encoding_size: 256
        learning_rate: 3.0e-4
      gail:
        gamma: 0.99
        strength: 0.05
        encoding_size: 128
        learning_rate: 3.0e-4
        use_actions: false
        use_vail: false
        demo_path: Demonstrations/
    behavioral_cloning:
      demo_path: Demonstrations/
      strength: 0.2
      steps: 2000000
      batch_size: 1024
      num_epoch: 3
    max_steps: 2000000
    time_horizon: 128
    summary_freq: 50000
    keep_checkpoints: 10
    checkpoint_interval: 100000
    threaded: false
    init_path: null
environment_parameters:
  min_size:
    curriculum:
      - name: Easy
        completion_criteria:
          measure: reward
          behavior: DungeonPlayer
          signal_smoothing: true
          min_lesson_length: 100
          threshold: 0.9
        value: 2.0
      - name: Medium
        value: 10.0
  max_size:
    curriculum:
      - name: Easy
        completion_criteria:
          measure: reward
          behavior: DungeonPlayer
          signal_smoothing: true
          min_lesson_length: 100
          threshold: 0.9
        value: 15.0
      - name: Medium
        value: 20.0
  walls:
    curriculum:
      - name: Easy
        completion_criteria:
          measure: reward
          behavior: DungeonPlayer
          signal_smoothing: true
          min_lesson_length: 100
          threshold: 0.9
        value: 0.1
      - name: Medium
        value: 0.2
  enemies:
    curriculum:
      - name: Easy
        completion_criteria:
          measure: reward
          behavior: DungeonPlayer
          signal_smoothing: true
          min_lesson_length: 100
          threshold: 0.9
        value: 0.0
      - name: Medium
        completion_criteria:
          measure: reward
          behavior: DungeonPlayer
          signal_smoothing: true
          min_lesson_length: 100
          threshold: 1.2
        value: 1.0
      - name: Hard
        value: 2.0
checkpoint_settings:
  run_id: dungeon_player
  initialize_from: null
  load_model: false
  resume: false
  force: false
  train_model: false
  inference: false
  results_dir: Results
env_settings:
  env_path: null
  env_args: null
  base_port: 5005
  num_envs: 1
  num_areas: 1
  timeout_wait: 60
  seed: -1
  max_lifetime_restarts: 10
  restarts_rate_limit_n: 1
  restarts_rate_limit_period_s: 60
engine_settings:
  width: 84
  height: 84
  quality_level: 5
  time_scale: 20
  target_frame_rate: -1
  capture_frame_rate: 60
  no_graphics: true
  no_graphics_monitor: false
torch_settings:
  device: null
debug: false