incorporated task parameter channel branch

added the ability to set task parameters from python

Project/Assets/ML-Agents/Examples/3DBall/TFModels/My3DBall.nn.meta

+fileFormatVersion: 2
+guid: 5fa1fb7f873d34cb5aad88303fef876d
+ScriptedImporter:
+  fileIDToRecycleName:
+    11400000: main obj
+    11400002: model data
+  externalObjects: {}
+  userData: 
+  assetBundleName: 
+  assetBundleVariant: 
+  script: {fileID: 11500000, guid: 19ed1486aa27d4903b34839f37b8f69f, type: 3}

config/ppo/RollerBall.yaml

+behaviors:
+  RollerBall:
+    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: 5
+      learning_rate_schedule: linear
+    network_settings:
+      normalize: true
+      hidden_units: 64
+      num_layers: 2
+      vis_encode_type: simple
+    reward_signals:
+      extrinsic:
+        gamma: 0.995
+        strength: 1.0
+    output_path: default
+    keep_checkpoints: 5
+    max_steps: 1000000
+    time_horizon: 1000
+    summary_freq: 12000
+    threaded: true

config/rollerball/rollerball_config.yaml

+behaviors:
+  RollerBall:
+    trainer: ppo
+    batch_size: 10
+    beta: 5.0e-3
+    buffer_size: 100
+    epsilon: 0.2
+    hidden_units: 128
+    lambd: 0.95
+    learning_rate: 3.0e-4
+    learning_rate_schedule: linear
+    max_steps: 5.0e4
+    memory_size: 128
+    normalize: false
+    num_epoch: 3
+    num_layers: 2
+    time_horizon: 64
+    summary_freq: 10000
+    use_recurrent: false
+    reward_signals:
+        extrinsic:
+            strength: 1.0
+            gamma: 0.99