beta...

config/trainer_config.yaml

     batch_size: 2048
     buffer_size: 20480
     hidden_units: 512
+    beta: 1.0e-2
     threaded: false
     time_horizon: 1000
     num_layers: 2
     batch_size: 2048
     buffer_size: 20480
     hidden_units: 512
+    beta: 1.0e-2
     threaded: false
     time_horizon: 1000
     num_layers: 2
     batch_size: 2048
     buffer_size: 20480
     hidden_units: 512
+    beta: 1.0e-2
     threaded: false
     time_horizon: 1000
     num_layers: 2