linear policy and linear forward

config/sac_transfer/3DBall.yaml

       save_replay_buffer: false
       init_entcoef: 0.5
       reward_signal_steps_per_update: 10.0
-      encoder_layers: 1
-      policy_layers: 1
-      forward_layers: 1
+      encoder_layers: 2
+      policy_layers: 0
+      forward_layers: 0
       value_layers: 1
       feature_size: 16
       # separate_value_net: true

config/sac_transfer/3DBallHard.yaml

       save_replay_buffer: false
       init_entcoef: 1.0
       reward_signal_steps_per_update: 10.0
-      encoder_layers: 1
-      policy_layers: 1
-      forward_layers: 1
+      encoder_layers: 2
+      policy_layers: 0
+      forward_layers: 0
       value_layers: 1
       feature_size: 16
       # separate_value_net: true

config/sac_transfer/3DBallHardTransfer.yaml

       save_replay_buffer: false
       init_entcoef: 1.0
       reward_signal_steps_per_update: 10.0
-      encoder_layers: 1
-      policy_layers: 1
-      forward_layers: 1
+      encoder_layers: 2
+      policy_layers: 0
+      forward_layers: 0
       value_layers: 1
       feature_size: 16
       # separate_value_net: true