benson V5 har cooket

MLAgents/config/BensonImitationV3-1env.yaml

@@ -0,0 +1,65 @@
+behaviors:
+  BensonImitationV3:
+    trainer_type: ppo
+
+    hyperparameters:
+      # Hyperparameters common to PPO and SAC
+      batch_size: 50
+      buffer_size: 3000
+      learning_rate: 3.0e-4
+      learning_rate_schedule: linear
+
+      # PPO-specific hyperparameters
+      # Replaces the "PPO-specific hyperparameters" section above
+      beta: 5.0e-2
+      epsilon: 0.1
+      lambd: 0.95
+      num_epoch: 3
+
+    # Configuration of the neural network (common to PPO/SAC)
+    network_settings:
+      vis_encoder_type: simple
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+
+    # Trainer configurations common to all trainers
+    max_steps: 1.8e5
+    time_horizon: 64
+    summary_freq: 1800
+    keep_checkpoints: 5
+    checkpoint_interval: 100000
+    threaded: true
+    init_path: null
+
+    # behavior cloning
+    behavioral_cloning:
+      demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+      strength: 0.5
+      # steps: 150000
+      # batch_size: 512
+      # num_epoch: 3
+      # samples_per_update: 0
+
+    reward_signals:
+      # environment reward (default)
+      extrinsic:
+        strength: 1.0
+        gamma: 0.99
+
+      # curiosity module
+      curiosity:
+        strength: 0.02
+        gamma: 0.99
+        encoding_size: 256
+        learning_rate: 3.0e-4
+
+      # GAIL
+      gail:
+        strength: 0.5
+        # gamma: 0.99
+        # encoding_size: 128
+        demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+        # learning_rate: 3.0e-4
+        # use_actions: false
+        # use_vail: false

MLAgents/config/BensonImitationV3-5env.yaml

@@ -0,0 +1,65 @@
+behaviors:
+  BensonImitationV3:
+    trainer_type: ppo
+
+    hyperparameters:
+      # Hyperparameters common to PPO and SAC
+      batch_size: 50
+      buffer_size: 15000
+      learning_rate: 3.0e-4
+      learning_rate_schedule: linear
+
+      # PPO-specific hyperparameters
+      # Replaces the "PPO-specific hyperparameters" section above
+      beta: 5.0e-2
+      epsilon: 0.1
+      lambd: 0.95
+      num_epoch: 3
+
+    # Configuration of the neural network (common to PPO/SAC)
+    network_settings:
+      vis_encoder_type: simple
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+
+    # Trainer configurations common to all trainers
+    max_steps: 2.4e5
+    time_horizon: 64
+    summary_freq: 9000
+    keep_checkpoints: 5
+    checkpoint_interval: 100000
+    threaded: true
+    init_path: null
+
+    # behavior cloning
+    behavioral_cloning:
+      demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+      strength: 0.5
+      # steps: 150000
+      # batch_size: 512
+      # num_epoch: 3
+      # samples_per_update: 0
+
+    reward_signals:
+      # environment reward (default)
+      extrinsic:
+        strength: 1.0
+        gamma: 0.99
+
+      # curiosity module
+      curiosity:
+        strength: 0.02
+        gamma: 0.99
+        encoding_size: 256
+        learning_rate: 3.0e-4
+
+      # GAIL
+      gail:
+        strength: 0.5
+        # gamma: 0.99
+        # encoding_size: 128
+        demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+        # learning_rate: 3.0e-4
+        # use_actions: false
+        # use_vail: false

MLAgents/config/BensonImitationV3.yaml

@@ -0,0 +1,65 @@
+behaviors:
+  BensonImitationV3:
+    trainer_type: ppo
+
+    hyperparameters:
+      # Hyperparameters common to PPO and SAC
+      batch_size: 50
+      buffer_size: 3000
+      learning_rate: 3.0e-4
+      learning_rate_schedule: linear
+
+      # PPO-specific hyperparameters
+      # Replaces the "PPO-specific hyperparameters" section above
+      beta: 5.0e-2
+      epsilon: 0.1
+      lambd: 0.95
+      num_epoch: 3
+
+    # Configuration of the neural network (common to PPO/SAC)
+    network_settings:
+      vis_encoder_type: simple
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+
+    # Trainer configurations common to all trainers
+    max_steps: 1.0e6
+    time_horizon: 64
+    summary_freq: 10000
+    keep_checkpoints: 5
+    checkpoint_interval: 100000
+    threaded: true
+    init_path: null
+
+    # behavior cloning
+    behavioral_cloning:
+      demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+      strength: 0.5
+      # steps: 150000
+      # batch_size: 512
+      # num_epoch: 3
+      # samples_per_update: 0
+
+    reward_signals:
+      # environment reward (default)
+      extrinsic:
+        strength: 1.0
+        gamma: 0.99
+
+      # curiosity module
+      curiosity:
+        strength: 0.02
+        gamma: 0.99
+        encoding_size: 256
+        learning_rate: 3.0e-4
+
+      # GAIL
+      gail:
+        strength: 0.5
+        # gamma: 0.99
+        # encoding_size: 128
+        demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+        # learning_rate: 3.0e-4
+        # use_actions: false
+        # use_vail: false

MLAgents/config/BensonV4.yaml

@@ -0,0 +1,65 @@
+behaviors:
+  BensonV4:
+    trainer_type: ppo
+
+    hyperparameters:
+      # Hyperparameters common to PPO and SAC
+      batch_size: 50
+      buffer_size: 10240
+      learning_rate: 3.0e-4
+      learning_rate_schedule: linear
+
+      # PPO-specific hyperparameters
+      # Replaces the "PPO-specific hyperparameters" section above
+      beta: 5.0e-2
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+
+    # Configuration of the neural network (common to PPO/SAC)
+    network_settings:
+      vis_encoder_type: simple
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+
+    # Trainer configurations common to all trainers
+    max_steps: 5.0e6
+    time_horizon: 64
+    summary_freq: 10000
+    keep_checkpoints: 5
+    checkpoint_interval : 500000
+    threaded: false
+    init_path: null
+
+    # # behavior cloning
+    # behavioral_cloning:
+    #   demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+    #   strength: 0.5
+    #   # steps: 150000
+    #   # batch_size: 512
+    #   # num_epoch: 3
+    #   # samples_per_update: 0
+
+    reward_signals:
+      # environment reward (default)
+      extrinsic:
+        strength: 1.0
+        gamma: 0.99
+
+      # # curiosity module
+      # curiosity:
+      #   strength: 0.02
+      #   gamma: 0.99
+      #   encoding_size: 256
+      #   learning_rate: 3.0e-4
+
+      # # GAIL
+      # gail:
+      #   strength: 0.5
+      #   # gamma: 0.99
+      #   # encoding_size: 128
+      #   demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+      #   # learning_rate: 3.0e-4
+      #   # use_actions: false
+      #   # use_vail: false

MLAgents/config/BensonV5.yaml

@@ -0,0 +1,65 @@
+behaviors:
+  BensonV5:
+    trainer_type: ppo
+
+    hyperparameters:
+      # Hyperparameters common to PPO and SAC
+      batch_size: 50
+      buffer_size: 10240
+      learning_rate: 3.0e-4
+      learning_rate_schedule: linear
+
+      # PPO-specific hyperparameters
+      # Replaces the "PPO-specific hyperparameters" section above
+      beta: 5.0e-2
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+
+    # Configuration of the neural network (common to PPO/SAC)
+    network_settings:
+      vis_encoder_type: simple
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+
+    # Trainer configurations common to all trainers
+    max_steps: 3.0e7
+    time_horizon: 64
+    summary_freq: 10000
+    keep_checkpoints: 5
+    checkpoint_interval : 500000
+    threaded: false
+    init_path: null
+
+    # # behavior cloning
+    # behavioral_cloning:
+    #   demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+    #   strength: 0.5
+    #   # steps: 150000
+    #   # batch_size: 512
+    #   # num_epoch: 3
+    #   # samples_per_update: 0
+
+    reward_signals:
+      # environment reward (default)
+      extrinsic:
+        strength: 1.0
+        gamma: 0.99
+
+      # # curiosity module
+      # curiosity:
+      #   strength: 0.02
+      #   gamma: 0.99
+      #   encoding_size: 256
+      #   learning_rate: 3.0e-4
+
+      # # GAIL
+      # gail:
+      #   strength: 0.5
+      #   # gamma: 0.99
+      #   # encoding_size: 128
+      #   demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+      #   # learning_rate: 3.0e-4
+      #   # use_actions: false
+      #   # use_vail: false

MLAgents/config/BensonV6.yaml

@@ -0,0 +1,65 @@
+behaviors:
+  BensonV5:
+    trainer_type: ppo
+
+    hyperparameters:
+      # Hyperparameters common to PPO and SAC
+      batch_size: 50
+      buffer_size: 10240
+      learning_rate: 3.0e-4
+      learning_rate_schedule: linear
+
+      # PPO-specific hyperparameters
+      # Replaces the "PPO-specific hyperparameters" section above
+      beta: 5.0e-2
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+
+    # Configuration of the neural network (common to PPO/SAC)
+    network_settings:
+      vis_encoder_type: simple
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+
+    # Trainer configurations common to all trainers
+    max_steps: 3.0e6
+    time_horizon: 64
+    summary_freq: 10000
+    keep_checkpoints: 5
+    checkpoint_interval : 500000
+    threaded: false
+    init_path: null
+
+    # # behavior cloning
+    # behavioral_cloning:
+    #   demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+    #   strength: 0.5
+    #   # steps: 150000
+    #   # batch_size: 512
+    #   # num_epoch: 3
+    #   # samples_per_update: 0
+
+    reward_signals:
+      # environment reward (default)
+      extrinsic:
+        strength: 1.0
+        gamma: 0.99
+
+      # # curiosity module
+      # curiosity:
+      #   strength: 0.02
+      #   gamma: 0.99
+      #   encoding_size: 256
+      #   learning_rate: 3.0e-4
+
+      # # GAIL
+      # gail:
+      #   strength: 0.5
+      #   # gamma: 0.99
+      #   # encoding_size: 128
+      #   demo_path: 'c:\Users\noahk\Documents\Unity projects\Racesm\Assets\Demonstrations\BensonV3M.demo'
+      #   # learning_rate: 3.0e-4
+      #   # use_actions: false
+      #   # use_vail: false

MLAgents/config/imitation/Crawler.yaml

@@ -0,0 +1,39 @@
+behaviors:
+  Crawler:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 2024
+      buffer_size: 20240
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: true
+      hidden_units: 512
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      gail:
+        gamma: 0.99
+        strength: 1.0
+        network_settings:
+          normalize: true
+          hidden_units: 128
+          num_layers: 2
+          vis_encode_type: simple
+        learning_rate: 0.0003
+        use_actions: false
+        use_vail: false
+        demo_path: Project/Assets/ML-Agents/Examples/Crawler/Demos/ExpertCrawler.demo
+    keep_checkpoints: 5
+    max_steps: 10000000
+    time_horizon: 1000
+    summary_freq: 30000
+    behavioral_cloning:
+      demo_path: Project/Assets/ML-Agents/Examples/Crawler/Demos/ExpertCrawler.demo
+      steps: 50000
+      strength: 0.5
+      samples_per_update: 0

MLAgents/config/imitation/Hallway.yaml

@@ -0,0 +1,35 @@
+behaviors:
+  Hallway:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 1024
+      learning_rate: 0.0003
+      beta: 0.01
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+      memory:
+        sequence_length: 64
+        memory_size: 256
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+      gail:
+        gamma: 0.99
+        strength: 0.01
+        learning_rate: 0.0003
+        use_actions: false
+        use_vail: false
+        demo_path: Project/Assets/ML-Agents/Examples/Hallway/Demos/ExpertHallway.demo
+    keep_checkpoints: 5
+    max_steps: 10000000
+    time_horizon: 64
+    summary_freq: 10000

MLAgents/config/imitation/PushBlock.yaml

@@ -0,0 +1,42 @@
+behaviors:
+  PushBlock:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.01
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+      gail:
+        gamma: 0.99
+        strength: 0.01
+        network_settings:
+          normalize: false
+          hidden_units: 128
+          num_layers: 2
+          vis_encode_type: simple
+        learning_rate: 0.0003
+        use_actions: false
+        use_vail: false
+        demo_path: Project/Assets/ML-Agents/Examples/PushBlock/Demos/ExpertPushBlock.demo
+    keep_checkpoints: 5
+    max_steps: 100000
+    time_horizon: 64
+    summary_freq: 60000
+    behavioral_cloning:
+      demo_path: Project/Assets/ML-Agents/Examples/PushBlock/Demos/ExpertPushBlock.demo
+      steps: 50000
+      strength: 1.0
+      samples_per_update: 0

MLAgents/config/imitation/Pyramids.yaml

@@ -0,0 +1,34 @@
+behaviors:
+  Pyramids:
+    trainer_type: ppo
+    time_horizon: 128
+    max_steps: 1.0e7
+    hyperparameters:
+      batch_size: 128
+      beta: 0.01
+      buffer_size: 2048
+      epsilon: 0.2
+      lambd: 0.95
+      learning_rate: 0.0003
+      num_epoch: 3
+    network_settings:
+      num_layers: 2
+      normalize: false
+      hidden_units: 512
+    reward_signals:
+      extrinsic:
+        strength: 1.0
+        gamma: 0.99
+      curiosity:
+        strength: 0.02
+        gamma: 0.99
+        network_settings:
+          hidden_units: 256
+      gail:
+        strength: 0.01
+        gamma: 0.99
+        demo_path: Project/Assets/ML-Agents/Examples/Pyramids/Demos/ExpertPyramid.demo
+    behavioral_cloning:
+      demo_path: Project/Assets/ML-Agents/Examples/Pyramids/Demos/ExpertPyramid.demo
+      strength: 0.5
+      steps: 150000

MLAgents/config/poca/DungeonEscape.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  DungeonEscape:
+    trainer_type: poca
+    hyperparameters:
+      batch_size: 1024
+      buffer_size: 10240
+      learning_rate: 0.0003
+      beta: 0.01
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: constant
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 20000000
+    time_horizon: 64
+    summary_freq: 60000

MLAgents/config/poca/PushBlockCollab.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  PushBlockCollab:
+    trainer_type: poca
+    hyperparameters:
+      batch_size: 1024
+      buffer_size: 10240
+      learning_rate: 0.0003
+      beta: 0.01
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: constant
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 15000000
+    time_horizon: 64
+    summary_freq: 60000

MLAgents/config/poca/SoccerTwos.yaml

@@ -0,0 +1,32 @@
+behaviors:
+  SoccerTwos:
+    trainer_type: poca
+    hyperparameters:
+      batch_size: 2048
+      buffer_size: 20480
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: constant
+    network_settings:
+      normalize: false
+      hidden_units: 512
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 50000000
+    time_horizon: 1000
+    summary_freq: 10000
+    self_play:
+      save_steps: 50000
+      team_change: 200000
+      swap_steps: 2000
+      window: 10
+      play_against_latest_model_ratio: 0.5
+      initial_elo: 1200.0

MLAgents/config/poca/StrikersVsGoalie.yaml

@@ -0,0 +1,63 @@
+behaviors:
+  Goalie:
+    trainer_type: poca
+    hyperparameters:
+      batch_size: 2048
+      buffer_size: 20480
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: constant
+    network_settings:
+      normalize: false
+      hidden_units: 512
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 30000000
+    time_horizon: 1000
+    summary_freq: 10000
+    self_play:
+      save_steps: 50000
+      team_change: 200000
+      swap_steps: 1000
+      window: 10
+      play_against_latest_model_ratio: 0.5
+      initial_elo: 1200.0
+  Striker:
+    trainer_type: poca
+    hyperparameters:
+      batch_size: 2048
+      buffer_size: 20480
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: constant
+    network_settings:
+      normalize: false
+      hidden_units: 512
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 30000000
+    time_horizon: 1000
+    summary_freq: 10000
+    self_play:
+      save_steps: 50000
+      team_change: 200000
+      swap_steps: 4000
+      window: 10
+      play_against_latest_model_ratio: 0.5
+      initial_elo: 1200.0

MLAgents/config/ppo/3DBall.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  3DBall:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 64
+      buffer_size: 12000
+      learning_rate: 0.0003
+      beta: 0.001
+      epsilon: 0.2
+      lambd: 0.99
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: true
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 500000
+    time_horizon: 1000
+    summary_freq: 12000

MLAgents/config/ppo/3DBallHard.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  3DBallHard:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 120
+      buffer_size: 12000
+      learning_rate: 0.0003
+      beta: 0.001
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: true
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 500000
+    time_horizon: 1000
+    summary_freq: 12000

MLAgents/config/ppo/3DBall_randomize.yaml

@@ -0,0 +1,36 @@
+behaviors:
+  3DBall:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 64
+      buffer_size: 12000
+      learning_rate: 0.0003
+      beta: 0.001
+      epsilon: 0.2
+      lambd: 0.99
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: true
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 500000
+    time_horizon: 1000
+    summary_freq: 12000
+environment_parameters:
+  mass:
+    sampler_type: uniform
+    sampler_parameters:
+      min_value: 0.5
+      max_value: 10
+  scale:
+    sampler_type: uniform
+    sampler_parameters:
+      min_value: 0.75
+      max_value: 3

MLAgents/config/ppo/Basic.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  Basic:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 32
+      buffer_size: 256
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 20
+      num_layers: 1
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.9
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 500000
+    time_horizon: 3
+    summary_freq: 2000

MLAgents/config/ppo/Crawler.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  Crawler:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 2048
+      buffer_size: 20480
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: true
+      hidden_units: 512
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.995
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 10000000
+    time_horizon: 1000
+    summary_freq: 30000

MLAgents/config/ppo/FoodCollector.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  GridFoodCollector:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 1024
+      buffer_size: 10240
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 1
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 2000000
+    time_horizon: 64
+    summary_freq: 10000

MLAgents/config/ppo/GridWorld.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  GridWorld:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 32
+      buffer_size: 256
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 128
+      num_layers: 1
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.9
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 500000
+    time_horizon: 5
+    summary_freq: 20000

MLAgents/config/ppo/Hallway.yaml

@@ -0,0 +1,28 @@
+behaviors:
+  Hallway:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 1024
+      learning_rate: 0.0003
+      beta: 0.03
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+      memory:
+        sequence_length: 64
+        memory_size: 128
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 10000000
+    time_horizon: 64
+    summary_freq: 10000

MLAgents/config/ppo/Match3.yaml

@@ -0,0 +1,48 @@
+default_settings:
+  trainer_type: ppo
+  hyperparameters:
+    batch_size: 16
+    buffer_size: 120
+    learning_rate: 0.0003
+    beta: 0.005
+    epsilon: 0.2
+    lambd: 0.99
+    num_epoch: 3
+    learning_rate_schedule: constant
+  network_settings:
+    normalize: true
+    hidden_units: 256
+    num_layers: 4
+    vis_encode_type: match3
+  reward_signals:
+    extrinsic:
+      gamma: 0.99
+      strength: 1.0
+  keep_checkpoints: 5
+  max_steps: 5000000
+  time_horizon: 128
+  summary_freq: 10000
+
+behaviors:
+  Match3SimpleHeuristic:
+    # Settings can be very simple since we don't care about actually training the model
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 16
+      buffer_size: 120
+    network_settings:
+      hidden_units: 4
+      num_layers: 1
+    max_steps: 5000000
+    summary_freq: 10000
+  Match3SmartHeuristic:
+    # Settings can be very simple since we don't care about actually training the model
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 16
+      buffer_size: 120
+    network_settings:
+      hidden_units: 4
+      num_layers: 1
+    max_steps: 5000000
+    summary_freq: 10000

MLAgents/config/ppo/PushBlock.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  PushBlock:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.01
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 2000000
+    time_horizon: 64
+    summary_freq: 60000

MLAgents/config/ppo/Pyramids.yaml

@@ -0,0 +1,31 @@
+behaviors:
+  Pyramids:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.01
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 512
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+      curiosity:
+        gamma: 0.99
+        strength: 0.02
+        network_settings:
+          hidden_units: 256
+        learning_rate: 0.0003
+    keep_checkpoints: 5
+    max_steps: 10000000
+    time_horizon: 128
+    summary_freq: 30000

MLAgents/config/ppo/PyramidsRND.yaml

@@ -0,0 +1,32 @@
+behaviors:
+  Pyramids:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.01
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 512
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+      rnd:
+        gamma: 0.99
+        strength: 0.01
+        network_settings:
+          hidden_units: 64
+          num_layers: 3
+        learning_rate: 0.0001
+    keep_checkpoints: 5
+    max_steps: 3000000
+    time_horizon: 128
+    summary_freq: 30000

MLAgents/config/ppo/Sorter_curriculum.yaml

@@ -0,0 +1,102 @@
+behaviors:
+  Sorter:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 512
+      buffer_size: 40960
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: constant
+    network_settings:
+      normalize: False
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 5000000
+    time_horizon: 256
+    summary_freq: 10000
+environment_parameters:
+  num_tiles:
+    curriculum:
+      - name: Lesson0 # The '-' is important as this is a list
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.3
+        value: 2.0
+      - name: Lesson1
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.4
+        value: 4.0
+      - name: Lesson2
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.45
+        value: 6.0
+      - name: Lesson3
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.5
+        value: 8.0
+      - name: Lesson4
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.55
+        value: 10.0
+      - name: Lesson5
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.6
+        value: 12.0
+      - name: Lesson6
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.65
+        value: 14.0
+      - name: Lesson7
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.7
+        value: 16.0
+      - name: Lesson8
+        completion_criteria:
+          measure: progress
+          behavior: Sorter
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.75
+        value: 18.0
+      - name: Lesson9
+        value: 20.0

MLAgents/config/ppo/Visual3DBall.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  Visual3DBall:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 256
+      buffer_size: 2560
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 400000
+    time_horizon: 64
+    summary_freq: 20000

MLAgents/config/ppo/VisualFoodCollector.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  VisualFoodCollector:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 1024
+      buffer_size: 10240
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 3000000
+    time_horizon: 100
+    summary_freq: 40000

MLAgents/config/ppo/Walker.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  Walker:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 2048
+      buffer_size: 20480
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: true
+      hidden_units: 512
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.995
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 30000000
+    time_horizon: 1000
+    summary_freq: 30000

MLAgents/config/ppo/WallJump.yaml

@@ -0,0 +1,49 @@
+behaviors:
+  BigWallJump:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 20000000
+    time_horizon: 128
+    summary_freq: 20000
+  SmallWallJump:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 5000000
+    time_horizon: 128
+    summary_freq: 20000

MLAgents/config/ppo/WallJump_curriculum.yaml

@@ -0,0 +1,118 @@
+behaviors:
+  BigWallJump:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 20000000
+    time_horizon: 128
+    summary_freq: 20000
+  SmallWallJump:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 128
+      buffer_size: 2048
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 5000000
+    time_horizon: 128
+    summary_freq: 20000
+environment_parameters:
+  big_wall_height:
+    curriculum:
+      - name: Lesson0 # The '-' is important as this is a list
+        completion_criteria:
+          measure: progress
+          behavior: BigWallJump
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.1
+        value:
+          sampler_type: uniform
+          sampler_parameters:
+            min_value: 0.0
+            max_value: 4.0
+      - name: Lesson1 # This is the start of the second lesson
+        completion_criteria:
+          measure: progress
+          behavior: BigWallJump
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.3
+        value:
+          sampler_type: uniform
+          sampler_parameters:
+            min_value: 4.0
+            max_value: 7.0
+      - name: Lesson2
+        completion_criteria:
+          measure: progress
+          behavior: BigWallJump
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.5
+        value:
+          sampler_type: uniform
+          sampler_parameters:
+            min_value: 6.0
+            max_value: 8.0
+      - name: Lesson3
+        value: 8.0
+  small_wall_height:
+    curriculum:
+      - name: Lesson0
+        completion_criteria:
+          measure: progress
+          behavior: SmallWallJump
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.1
+        value: 1.5
+      - name: Lesson1
+        completion_criteria:
+          measure: progress
+          behavior: SmallWallJump
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.3
+        value: 2.0
+      - name: Lesson2
+        completion_criteria:
+          measure: progress
+          behavior: SmallWallJump
+          signal_smoothing: true
+          min_lesson_length: 100
+          threshold: 0.5
+        value: 2.5
+      - name: Lesson3
+        value: 4.0

MLAgents/config/ppo/Worm.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  Worm:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 2024
+      buffer_size: 20240
+      learning_rate: 0.0003
+      beta: 0.005
+      epsilon: 0.2
+      lambd: 0.95
+      num_epoch: 3
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: true
+      hidden_units: 512
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.995
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 7000000
+    time_horizon: 1000
+    summary_freq: 30000

MLAgents/config/rollerball_config.yaml

@@ -0,0 +1,25 @@
+behaviors:
+  RollerBall:
+    trainer_type: ppo
+    hyperparameters:
+      batch_size: 10
+      buffer_size: 100
+      learning_rate: 3.0e-4
+      beta: 5.0e-4
+      epsilon: 0.2
+      lambd: 0.99
+      num_epoch: 3
+      learning_rate_schedule: linear
+      beta_schedule: constant
+      epsilon_schedule: linear
+    network_settings:
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    max_steps: 500000
+    time_horizon: 64
+    summary_freq: 10000

MLAgents/config/sac/3DBall.yaml

@@ -0,0 +1,27 @@
+behaviors:
+  3DBall:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 64
+      buffer_size: 200000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 10.0
+      save_replay_buffer: false
+      init_entcoef: 0.5
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: true
+      hidden_units: 64
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 200000
+    time_horizon: 1000
+    summary_freq: 12000

MLAgents/config/sac/3DBallHard.yaml

@@ -0,0 +1,27 @@
+behaviors:
+  3DBallHard:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 256
+      buffer_size: 500000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 10.0
+      save_replay_buffer: false
+      init_entcoef: 1.0
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: true
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 500000
+    time_horizon: 1000
+    summary_freq: 12000

MLAgents/config/sac/Basic.yaml

@@ -0,0 +1,27 @@
+behaviors:
+  Basic:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 64
+      buffer_size: 50000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 10.0
+      save_replay_buffer: false
+      init_entcoef: 0.01
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: false
+      hidden_units: 20
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 500000
+    time_horizon: 10
+    summary_freq: 2000

MLAgents/config/sac/Crawler.yaml

@@ -0,0 +1,27 @@
+behaviors:
+  Crawler:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 256
+      buffer_size: 500000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 20.0
+      save_replay_buffer: false
+      init_entcoef: 1.0
+      reward_signal_steps_per_update: 20.0
+    network_settings:
+      normalize: true
+      hidden_units: 512
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.995
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 5000000
+    time_horizon: 1000
+    summary_freq: 30000

MLAgents/config/sac/FoodCollector.yaml

@@ -0,0 +1,28 @@
+behaviors:
+  FoodCollector:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 256
+      buffer_size: 2048
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 10.0
+      save_replay_buffer: false
+      init_entcoef: 0.05
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 1
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 2000000
+    time_horizon: 64
+    summary_freq: 60000
+    threaded: false

MLAgents/config/sac/GridWorld.yaml

@@ -0,0 +1,27 @@
+behaviors:
+  GridWorld:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 128
+      buffer_size: 50000
+      buffer_init_steps: 1000
+      tau: 0.005
+      steps_per_update: 10.0
+      save_replay_buffer: false
+      init_entcoef: 0.5
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: false
+      hidden_units: 128
+      num_layers: 1
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.9
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 500000
+    time_horizon: 5
+    summary_freq: 20000

MLAgents/config/sac/Hallway.yaml

@@ -0,0 +1,30 @@
+behaviors:
+  Hallway:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 512
+      buffer_size: 200000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 10.0
+      save_replay_buffer: false
+      init_entcoef: 0.1
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: false
+      hidden_units: 128
+      num_layers: 2
+      vis_encode_type: simple
+      memory:
+        sequence_length: 64
+        memory_size: 128
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 4000000
+    time_horizon: 64
+    summary_freq: 10000

MLAgents/config/sac/PushBlock.yaml

@@ -0,0 +1,27 @@
+behaviors:
+  PushBlock:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 128
+      buffer_size: 50000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 10.0
+      save_replay_buffer: false
+      init_entcoef: 0.05
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 2000000
+    time_horizon: 64
+    summary_freq: 100000

MLAgents/config/sac/Pyramids.yaml

@@ -0,0 +1,34 @@
+behaviors:
+  Pyramids:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 128
+      buffer_size: 2000000
+      buffer_init_steps: 1000
+      tau: 0.01
+      steps_per_update: 10.0
+      save_replay_buffer: false
+      init_entcoef: 0.01
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: false
+      hidden_units: 512
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.995
+        strength: 2.0
+      gail:
+        gamma: 0.99
+        strength: 0.01
+        learning_rate: 0.0003
+        use_actions: true
+        use_vail: false
+        demo_path: Project/Assets/ML-Agents/Examples/Pyramids/Demos/ExpertPyramid.demo
+    keep_checkpoints: 5
+    max_steps: 3000000
+    time_horizon: 128
+    summary_freq: 30000

MLAgents/config/sac/Walker.yaml

@@ -0,0 +1,27 @@
+behaviors:
+  Walker:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 1024
+      buffer_size: 2000000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 30.0
+      save_replay_buffer: false
+      init_entcoef: 1.0
+      reward_signal_steps_per_update: 30.0
+    network_settings:
+      normalize: true
+      hidden_units: 256
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.995
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 15000000
+    time_horizon: 1000
+    summary_freq: 30000

MLAgents/config/sac/WallJump.yaml

@@ -0,0 +1,53 @@
+behaviors:
+  BigWallJump:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 128
+      buffer_size: 200000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 20.0
+      save_replay_buffer: false
+      init_entcoef: 0.1
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 15000000
+    time_horizon: 128
+    summary_freq: 20000
+  SmallWallJump:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 128
+      buffer_size: 50000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 20.0
+      save_replay_buffer: false
+      init_entcoef: 0.1
+      reward_signal_steps_per_update: 10.0
+    network_settings:
+      normalize: false
+      hidden_units: 256
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 5000000
+    time_horizon: 128
+    summary_freq: 20000

MLAgents/config/sac/Worm.yaml

@@ -0,0 +1,27 @@
+behaviors:
+  Worm:
+    trainer_type: sac
+    hyperparameters:
+      learning_rate: 0.0003
+      learning_rate_schedule: constant
+      batch_size: 256
+      buffer_size: 500000
+      buffer_init_steps: 0
+      tau: 0.005
+      steps_per_update: 20.0
+      save_replay_buffer: false
+      init_entcoef: 1.0
+      reward_signal_steps_per_update: 20.0
+    network_settings:
+      normalize: true
+      hidden_units: 512
+      num_layers: 3
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.995
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 5000000
+    time_horizon: 1000
+    summary_freq: 30000