Add PushBlockCollab config and fix some stuff

ml-agents/mlagents/trainers/coma/trainer.py

         )
         self.seed = seed
         self.policy: Policy = None  # type: ignore
-        self.collected_rewards["environment_team"] = defaultdict(lambda: 0)
+        self.collected_group_rewards = defaultdict(lambda: 0)
 
     def _process_trajectory(self, trajectory: Trajectory) -> None:
         """
         self.collected_rewards["environment"][agent_id] += np.sum(
             agent_buffer_trajectory[BufferKey.ENVIRONMENT_REWARDS]
         )
-        self.collected_rewards["environment_team"][agent_id] += np.sum(
+        self.collected_group_rewards[agent_id] += np.sum(
             agent_buffer_trajectory[BufferKey.GROUP_REWARD]
         )
         for name, reward_signal in self.optimizer.reward_signals.items():
         if "environment_team" in self.collected_rewards:
             self.stats_reporter.add_stat(
                 "Environment/Team Cumulative Reward",
-                self.collected_rewards["environment_team"].get(agent_id, 0),
+                self.collected_group_rewards.get(agent_id, 0),
                 aggregation=StatsAggregationMethod.HISTOGRAM,
             )
 

ml-agents/mlagents/trainers/settings.py

     def to_settings(self) -> type:
         _mapping = {
             RewardSignalType.EXTRINSIC: RewardSignalSettings,
+            RewardSignalType.GROUP_EXTRINSIC: RewardSignalSettings,
             RewardSignalType.GAIL: GAILSettings,
             RewardSignalType.CURIOSITY: CuriositySettings,
             RewardSignalType.RND: RNDSettings,

config/ppo/PushBlockCollab.yaml

+behaviors:
+  PushBlock:
+    trainer_type: coma
+    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:
+      group:
+        gamma: 0.99
+        strength: 1.0
+    keep_checkpoints: 5
+    max_steps: 20000000 #2000000
+    time_horizon: 64
+    summary_freq: 60000
+    threaded: true