no target net

config/ppo/PushBlock.yaml

       batch_size: 128
       buffer_size: 2048
       learning_rate: 0.0003
-      beta: 0.01
+      beta: 0.005
-      lambd: 0.95
+      lambd: 0.8
       num_epoch: 3
       learning_rate_schedule: linear
     network_settings:

ml-agents/mlagents/trainers/optimizer/torch_optimizer.py

 
         memory = torch.zeros([1, 1, self.policy.m_size])
 
-        value_estimates, marg_val_estimates, mem = self.policy.actor_critic.target_critic_pass(
+        value_estimates, marg_val_estimates, mem = self.policy.actor_critic.critic_pass(
             current_obs,
             actions,
             memory,
         )
-        next_value_estimates, next_marg_val_estimates, next_mem = self.policy.actor_critic.target_critic_pass(
+        next_value_estimates, next_marg_val_estimates, next_mem = self.policy.actor_critic.critic_pass(
             next_obs,
             next_actions,
             memory,

ml-agents/mlagents/trainers/ppo/optimizer_torch.py

         self.optimizer.step()
 
         ModelUtils.soft_update(
-            self.policy.actor_critic.critic, self.policy.actor_critic.target, .001
+            self.policy.actor_critic.critic, self.policy.actor_critic.target, 1.
         )
 
         update_stats = {