all self-play

config/trainer_config.yaml

     batch_size: 2048
     buffer_size: 20480
     hidden_units: 512
-    beta: 1.0e-2
-        window: 100
+        window: 10
         play_against_latest_model_ratio: 0.5
         save_steps: 50000
         swap_steps: 2500
     batch_size: 2048
     buffer_size: 20480
     hidden_units: 512
-    beta: 1.0e-2
-        window: 100
+        window: 10
         play_against_latest_model_ratio: 0.5
         save_steps: 50000
         swap_steps: 10000
     batch_size: 2048
     buffer_size: 20480
     hidden_units: 512
-    beta: 1.0e-2
-        window: 100
+        window: 10
         play_against_latest_model_ratio: 0.5
         save_steps: 50000
         swap_steps: 2000
     batch_size: 2048
     buffer_size: 20480
     hidden_units: 512
-    beta: 1.0e-2
     time_horizon: 1000
     num_layers: 2
     self_play:
     batch_size: 2048
     buffer_size: 20480
     hidden_units: 512
-    beta: 1.0e-2
     time_horizon: 1000
     num_layers: 2
     self_play: